BEGIN:VCALENDAR VERSION:2.0 PRODID:-//HasGeek//NONSGML Funnel//EN DESCRIPTION:Maximising the Potential of Data — Discussions around data s cience\, machine learning & AI X-WR-CALDESC:Maximising the Potential of Data — Discussions around data science\, machine learning & AI NAME:The Fifth Elephant 2024 Annual Conference (12th &13th July) X-WR-CALNAME:The Fifth Elephant 2024 Annual Conference (12th &13th July) REFRESH-INTERVAL;VALUE=DURATION:PT12H SUMMARY:The Fifth Elephant 2024 Annual Conference (12th &13th July) TIMEZONE-ID:Asia/Kolkata X-PUBLISHED-TTL:PT12H X-WR-TIMEZONE:Asia/Kolkata BEGIN:VEVENT SUMMARY:Check-in and registrations DTSTART:20240713T033000Z DTEND:20240713T040500Z DTSTAMP:20260419T104847Z UID:session/Wz3KS3Pj9uXVL25CkcpipH@hasgeek.com SEQUENCE:3 CREATED:20240625T100156Z LAST-MODIFIED:20240625T100834Z LOCATION:Bangalore & Hybrid ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Check-in and registrations in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Introduction to GenAI and LLMs track\; how to make the most of it DTSTART:20240713T040500Z DTEND:20240713T041500Z DTSTAMP:20260419T104847Z UID:session/4g6HWJgwE2x8iUgmUdPkeB@hasgeek.com SEQUENCE:2 CREATED:20240625T104555Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240625T104601Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Introduction to GenAI and LLMs track\; how to make the most of it in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Introduction to the Data Engineering and Infrastructure track - ho w to make the most of it DTSTART:20240713T040500Z DTEND:20240713T041500Z DTSTAMP:20260419T104847Z UID:session/4ykp7ZwZVR9YbvzSkSErvB@hasgeek.com SEQUENCE:2 CREATED:20240625T100232Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240625T100837Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Introduction to the Data Engineering and Infrastructure track - how to make the most of it in Data engineering and infrastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Power up LLMs to be Factual Ninjas using GraphRAGs DTSTART:20240713T041500Z DTEND:20240713T045000Z DTSTAMP:20260419T104847Z UID:session/W2Ta7KCR69viSfdXfwoXJB@hasgeek.com SEQUENCE:10 CATEGORIES:30 mins talk CREATED:20240625T104325Z DESCRIPTION:In the era of big data\, large language models (LLMs) are beco ming increasingly important for tasks like question answering\, document a nalysis\, and chatbot development. However\, traditional LLMs can often st ruggle with factual accuracy\, reasoning\, and handling complex informatio n.\n\nThis talk introduces GraphRAG\, a novel technique that leverages the power of knowledge graphs to empower LLMs. We'll delve into the limitatio ns of traditional LLMs and vector-only approach of RAG and explore how Gra phRAG bridges the gap. GraphRAG equips LLMs with a deeper understanding of the world by utilizing enterprise data\, boosting their accuracy\, reason ing abilities\, and handling of complex information.\n\nThis talk will exp lore the exciting world of GraphRAG\, showcasing its potential to revoluti onize various fields. We will also discuss real-world applications and cas e studies using SEC EDGAR dataset to illustrate the benefits of GraphRAG f or practitioners in comparison to a vector-only approach.\n\n1. Who is the audience for your session?\n- Machine learning engineers and researchers interested in large language models (LLMs)\n- Developers who want to build more advanced chatbots and virtual assistants\n- Anyone interested in the future of AI and how LLMs are being improved\n \n2. What problem/pain ar e you trying to solve (for the audience)?\n- Traditional LLMs can struggle with factual accuracy\, reasoning\, and handling complex information.\n- This can lead to unreliable chatbots\, inaccurate answers to questions\, a nd difficulty understanding the nuances of human language.\n \n3. What wi ll be the scope of your session?\n- This talk will introduce GraphRAG\, a new technique that uses knowledge graphs to improve the performance of LLM s.\n- We will discuss the limitations of traditional LLMs and how GraphRAG addresses these limitations.\n- We will also explore the potential applic ations of GraphRAG in various fields\, such as question answering\, docume nt analysis\, and chatbot development.\n \n4. How will participants benef it from your session?\n- Participants will gain a better understanding of how LLMs work and the challenges they face.\n- They will learn about Graph RAG and how it can be used to improve the accuracy and reasoning abilities of LLMs.\n- Participants will also walk away with practical tips on how t o use GraphRAG in their own work.\n \n5. What is the appropriate format f or your session\, given the audience and objectives that you have in mind? \n- This session will be a talk\, with a combination of lecture and presen tation slides.\n- There will also be a brief demonstration of GraphRAG in action. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071233Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/graphrag-powering-up-l lms-with-knowledge-graphs-W2Ta7KCR69viSfdXfwoXJB BEGIN:VALARM ACTION:display DESCRIPTION:Power up LLMs to be Factual Ninjas using GraphRAGs in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Data extraction from cloud using Atlassian Lithium Platform DTSTART:20240713T041500Z DTEND:20240713T045000Z DTSTAMP:20260419T104847Z UID:session/AwW4MvFDcLcGYRpN1NaoX6@hasgeek.com SEQUENCE:15 CATEGORIES:30 mins talk CREATED:20240625T101636Z DESCRIPTION:There are numerous use cases that require moving large amounts of data between different systems and validating and transforming them in -flight. Platforms such as Apache Flink can be excellent choices for movin g and transforming data at scale - effectively through streaming ETL. Howe ver\, certain use cases within Atlassian ‘onprem to cloud data migration '\, ‘cloud to cloud data migration’\, ‘backup and restore’\, and ' sandbox’ present unique challenges. They require stream processing pipel ines to be entirely provisioned at runtime\, including dedicated Kafka top ics\, parallelism\, selection of stream processors with appropriate and av ailable compute resources. Stream processors must be embedded directly in product and platform services to enable in-process access to service conte xt in order to meet functional requirements along with throughput goals. I n addition\, our pipelines require coordination amongst sources\, transfor ms and sinks that exist in different product and infrastructure services. This led us to build the Lithium Platform. It is 100% event driven and is built on Kafka and Kafka Streams.\n\nLithium is a platform that leverages Kafka Streaming technology based ETL pipelines to enable rapid and scalabl e data movement. It is designed for seamlessly transferring data between d ifferent points with maximum efficiency and reliability. It boasts of adva nced capabilities such as decoupled ETL components\, highly customizable j ob configuration options\, in-progress remediation capabilities\, continuo us job monitoring\, etc.\n\n## Outline\n\n### Introduction to Lithium Plat form\n\nWe will introduce the Lithium platform and provide an overview of why it was built in Atlassian and what kind of use cases it is solving cur rently. Introduce to the concept of ephemeral data pipelines\, a.k.a Workp lans\n\n### Lithium Control plane\n\nIntroduce the Lithium control plane t o the audience and discuss various components within Control Plane and wha t is it responsible for in managing Data Pipelines\n\n### Lithium Data Pla ne\n\nLithium Data Plane consists of the components responsible for moving the data through the data pipeline. We will go over various components of a Data Plane. We will also cover how we have designed Self Hosted Data Pl ane components\, OR BYOH (Bring your own host) with the help of Lithium SD Ks\n\n### Control plane and data plane communication\n\nWe will discuss ho w the Control plane and Data Plane communicates with each other. What kind s to events are exchanged between these systems to keep the Data Pipelines up and running\n\n### Custom source\, transform\, validation\, sink\, and workplan state processors\n\nLithium provides simple APIs through the dat a plane SDKs to write custom data plane processors (extract\, transform\, load etc). These processors together make the Data Plane and run the custo m logic within the Data Pipelines.\n\n### Dynamic Resource Provisioning in the Data Plane and the Auctioning Model\n\nLithium Data plane implements an Auctioning model to provision compute resources required for a particul ar job and all the resources\, including the kafka topics\, are provisione d on-demand. We will glance through how this has been implemented in Lithi um.\n\n### Data pipeline pause\, resume\, rewind\, and in-flight data reme diation via sidelining\n\nWe will discuss how we can pause\, resume and re wind data pipelines by issuing commands to the control plane. We will also discuss about Lithium's Sidelining capabilities and how it helps with in- flight remediation in the pipeline.\n\n## Impact\n{Replace this with an ex planation of the impact of your work within your organization.} GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071249Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/ephemeral-data-pipelin es-using-atlassians-lithium-platform-AwW4MvFDcLcGYRpN1NaoX6 BEGIN:VALARM ACTION:display DESCRIPTION:Data extraction from cloud using Atlassian Lithium Platform in Data engineering and infrastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T045000Z DTEND:20240713T050000Z DTSTAMP:20260419T104847Z UID:session/AB9zTdeQcRMCjexSaU4V6g@hasgeek.com SEQUENCE:4 CREATED:20240625T104611Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054228Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in LLMs and GenAI track (Auditorium on the ground f loor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T045000Z DTEND:20240713T050000Z DTSTAMP:20260419T104847Z UID:session/LKaxsNadkVPViVpuih2x5E@hasgeek.com SEQUENCE:4 CREATED:20240625T100645Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054225Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in Data engineering and infrastructure track (Semin ar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Multimodal Fusion: Fusing Diverse Data Points for Combating Comple x Challenges DTSTART:20240713T050000Z DTEND:20240713T053500Z DTSTAMP:20260419T104847Z UID:session/Vw2NN1jpc4hFcQruWLmuCQ@hasgeek.com SEQUENCE:10 CATEGORIES:30 mins talk CREATED:20240625T104643Z DESCRIPTION:Multimodal fusion has revolutionized the way we integrate and interpret diverse data sources\, creating powerful insights from the syner gy of visual\, auditory\, and textual information. In this talk\, titled " Unifying Senses: The Evolution\, Technology\, and Impact of Multimodal Fus ion\," we will explore the origins of multimodal fusion and trace its deve lopment over the years. We'll delve into how this technology has evolved t o solve complex problems by combining multiple data sources for enhanced i nsights.\n\nWe'll discuss the necessary infrastructure for developing soph isticated multimodal models and the advantages of deploying them on edge d evices for real-time applications. Furthermore\, we'll highlight the impor tant applications of multimodal fusion in the space industry\, addressing critical issues such as climate change\, environmental sustainability\, an d defense. Join us to discover how multimodal fusion is not only transform ing technology but also contributing to solving some of the most pressing challenges of our time.\n\n## Outline\n1. The origins - Combining the sens es.\n2. A trip down memory lane - Early experiments\n3. The rise of AI and machine learning - Enhancing fusion capabilities.\n4. How multimodal fusi on changed the game - Transformative moments.\n5. Real-world stories - Ind ustries and lives transformed.\n9. The nuts and bolts - Infrastructure beh ind multimodal fusion.\n10. Building powerful models - The blend of hardwa re and software.\n11. Edge computing - Bringing fusion capabilities closer to the user.\n12. Space and beyond - How multimodal fusion is tackling cl imate change.\n13. The future - Pioneering sustainable solutions for a bet ter planet.\n\n \n## Impact\nMultimodal fusion is crucial for developing effective strategies to combat climate change\, enable sustainibility at scale and support effective monitoring. At GalaxEye Space\, we are working to combine data frim world's first hybrid Earth observation sensors\, usi ng Multimodal fusion. As we continue to explore the cosmos\, we are workin g towards creating technology to harness the power of real time Analytics (ML\, DL)\, directly from Space.\n\n### Audience for the session\nWorking professionals in field of statistics\, ai & machine learning. Academicians and research scholars who want to explore applied research and trends in industry.\n\n### Scope & Problem to solve\n1. Understanding of Multimodal Fusion- Gain insights into the concept of multimodal fusion and how it int egrates visual\, auditory\, and textual information to create powerful ins ights.\n\n2. Evolution of Technology - Explore the historical progression of multimodal fusion technology\, from its origins to its current state\, and understand the key milestones that have shaped its development.\n\n3. Practical Applications - Learn about the diverse applications of multimod al fusion technology\, particularly in the space industry\, including its role in addressing critical issues such as climate change\, environmental sustainability\, and defense.\n\n4. Infrastructure Requirements & Tec stac k- Discover the necessary infrastructure for developing sophisticated mult imodal models and understand the advantages of deploying them on edge devi ces for real-time applications.\n\n5. Impact on Problem Solving- Gain an appreciation for how multimodal fusion is revolutionizing problem-solving by enabling the combination of multiple data sources to provide enhanced i nsights\, leading to more effective solutions for complex challenges.\n\n# ## Benefits\n1. Hands-on demonstration.\n2. Industrial usecases and applic ations.\n3. Exposure to current trends.\n4. opportunity to work with the o rganisation.\n5. Mentorship for selected candidates.\n\n\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071110Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/unifying-senses-the-ev olution-technology-and-impact-of-multimodal-fusion-Vw2NN1jpc4hFcQruWLmuCQ BEGIN:VALARM ACTION:display DESCRIPTION:Multimodal Fusion: Fusing Diverse Data Points for Combating Co mplex Challenges in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Vector databases: a bird's eye view DTSTART:20240713T050000Z DTEND:20240713T053500Z DTSTAMP:20260419T104847Z UID:session/Tju7RRLQCE5XekLCqf8c4q@hasgeek.com SEQUENCE:12 CATEGORIES:30 mins talk CREATED:20240625T100754Z DESCRIPTION:This talk is focused on an equipping the audience with an over all understanding of the current vector database landscape\, and how vecto r databases work internally with a focus on a few common algorithms. \n\n# # Audience\nPrimarily intended for engineers looking to understand the int ersection of information retrieval with artificial intelligence.\n\n## Sco pe\n1. Algorithms to store and retrieve vector embeddings are not a recent development. However\, in a growing market for vector DBs\, they are comi ng to the forefront with databases competing to build competitive vector i ndexing and retrieval capabilities.\n2. What is driving this shift to a ne w paradigm and why?\n3. Explore why established commonly used indexing alg orithms are not up to the task.\n4. Understand some of the popular algorit hms that address these limitations and the vector indexing offerings that have adopted these.\n5. Evaluate the capabilities offered by custom-built\ , dedicated vector databases vs those by integrating vector search into an existing database.\n\n## Takeaway\n1. Understanding of the current vector database landscape. \n2. Equip the audience with an understanding of the broad classification of Approximate Nearest Neighbour search algorithms.\n 3. The above\, in turn\, is a foundation to understand the fundamental alg orithms behind vector indexing and search.\n\n## Pain Points\nSince soluti ons to store and retrieve vector embeddings are now a key infrastructural component as part of RAG pipelines\, evaluating the various offerings will require understanding the most common algorithms.\n\nThis involves unders tanding of some aspects of multiple moving parts such as tuning embedding models\, leveraging the embedding to provide context in RAG pipelines.\n\n This talk aims to provide a system overview to build upon when surveying t he developments in this domain. \n\n## Format\nThe narrative will start fr om why vector databases are needed and the recent developments driving thi s demand. It will then transition into discussing the internals of vector databases and the specific trade-offs to consider when evaluating vector d atabases compared to non-vector ones.\n\n\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T082153Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/vector-databases-a-bir ds-eye-view-Tju7RRLQCE5XekLCqf8c4q BEGIN:VALARM ACTION:display DESCRIPTION:Vector databases: a bird's eye view in Data engineering and in frastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Sponsored Round Table - AI and its impact on developer productivit y DTSTART:20240713T053500Z DTEND:20240713T070000Z DTSTAMP:20260419T104847Z UID:session/N2pty4JzTr3b2Ae8Ve8o46@hasgeek.com SEQUENCE:7 CREATED:20240701T100525Z DESCRIPTION:AI has revolutionized various aspects of our daily routines\, enhancing efficiency\, decision-making\, and problem-solving skills. The c ore focus of AI lies in boosting productivity and seamlessly integrating i t into our day-to-day tasks. Our discussion will center around developers\ , delving into how AI is utilized by organizations to amplify developer pr oductivity.\n\n**Key discussion points:**\n\n1. Problem scope: understandi ng the challenges and solutions related to developer productivity in diffe rent organizations.\n2. Applications: exploring the use of familiar techno logies to enhance developer productivity.\n3. Networking opportunities: co nnecting with individuals sharing an interest in enhancing developer produ ctivity\, fostering potential collaborations.\n4. Inspiration and motivati on: showcasing real-world projects and success stories that illustrate the power and benefits of well-executed developer productivity initiatives.\n \nIn conclusion\, the roundtable discussion on AI's impact on developer pr oductivity is poised to offer an enlightening forum for delving into innov ative solutions aimed at improving efficiency and problem-solving abilitie s. By examining the challenges and uses of AI in enhancing developer produ ctivity\, participants can share valuable insights and success narratives that underscore the significant impact of well-planned endeavors in this f ield. We invite you to participate in this stimulating dialogue and discov er the possibilities of AI in optimizing developer productivity. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T055225Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Sponsored Round Table - AI and its impact on developer product ivity in Board room (second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Morning break DTSTART:20240713T053500Z DTEND:20240713T060500Z DTSTAMP:20260419T104847Z UID:session/SauVoJLtHkHUf8YH4Jfc3E@hasgeek.com SEQUENCE:9 CREATED:20240625T104701Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054248Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Morning break in LLMs and GenAI track (Auditorium on the groun d floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Morning break DTSTART:20240713T053500Z DTEND:20240713T060500Z DTSTAMP:20260419T104847Z UID:session/KJDB25LWwE3C5ZspMk8L7z@hasgeek.com SEQUENCE:6 CREATED:20240625T100819Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054244Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Morning break in Data engineering and infrastructure track (Se minar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Unfreeze your data markets with privacy-preserving DPI DTSTART:20240713T060000Z DTEND:20240713T063500Z DTSTAMP:20260419T104847Z UID:session/9uwV4gB1J99ZUSZvj7jEau@hasgeek.com SEQUENCE:9 CATEGORIES:30 mins talk CREATED:20240625T113530Z DESCRIPTION:Clearly\, there is a race to build larger and larger AI models these days trained on as much training data as possible. Indian developer s are also trying to make their presence felt in this race to build home-g rown models for our unique problems and situations. Owing to our large pop ulation\, we probably generate more data than any other country. This soun ds great\, right\, but much of this Indian context data is not in one plac e and is hidden behind numerous government and corporate walls. What makes the situation worse is most of these data silos are enterprises of tradit ional nature and are not the typical centers of innovation\, at least for modern technologies like AI. \nThis is a fertile ground for DPI (Digital P ublic Infrastructure) for AI . The three core concepts (Confidential Compu ting\, Differential Privacy and Smart Contracts) of DPI for AI ensure that this data sitting in silos can be seamlessly and democratically shared wi th innovators around India in a privacy-preserving manner. The techno-lega l framework makes it super easy for anyone to abide by the privacy regulat ions without sweat. This unfreezing of data is critical for our innovators to get easy access to contextual data to give them a much-needed leg up a gainst the western onslaught in the field of AI. This is India’s chance to leapfrog in the field of AI as we have done in so many domains (payment s\, identity\, internet\, etc.).\n\n## Outline\n- DPI for AI as a credible solution to increasing data needs for AI model development\n- Overall arc hitecture of the training architecture\n- Deep dive into concepts from Dif ferential Privacy (DP) which is a core pillar that enables safe data colla boration \n## Impact\n- Understand privacy as part of the design construct instead of an afterthought when we train models\n- Get acclimitised to a new DPI which is currently getting rolled out and get a chance to become a n early adopter for the same\n \n## Who is the audience for this? \n- AI practitioners who are wondering where the data will come from - particular ly folks in startups.\n- Developers/product folks who build models using consumer data. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240719T113807Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/a-privacy-preserving-d pi-to-unfreeze-data-markets-to-solve-our-data-woes-9uwV4gB1J99ZUSZvj7jEau BEGIN:VALARM ACTION:display DESCRIPTION:Unfreeze your data markets with privacy-preserving DPI in Work shop\; BOFs for LLMs and GenAI track (Library on the second floor) in 5 mi nutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:From foundation to the future: evolution of Dream11's data platfor m DTSTART:20240713T060500Z DTEND:20240713T064000Z DTSTAMP:20260419T104847Z UID:session/VMe4LzYi8T4fkaCjBw9T1i@hasgeek.com SEQUENCE:10 CATEGORIES:30 mins talk CREATED:20240628T050527Z DESCRIPTION:\n### Outline\n* Introduction (5 minutes)\n* Brief overview of Dream11\n* Importance of a robust data platform in fantasy sports\n\nInit ial Data Platform Setup (5 minutes)\nEarly challenges and requirements\nBa sic infrastructure and tools used\n\nScaling Up (5 minutes)\nGrowth in use r base and data volume\nIntroduction of new technologies and architectures \nKey challenges faced during scaling\n\n\nModern Data Ecosystem (10 minut es)\nCurrent architecture and technologies\nData pipelines\, storage solut ions\, and processing frameworks\nReal-time analytics and machine learning integration\n\n\nLessons Learned and Best Practices (3 minutes)\nKey take aways from the evolution process\nBest practices for building and scaling a data platform\n\n\n### Q&A (2 minutes)\n\nAudience for the session\nThe primary audience for this session includes data engineers\, data architect s\, and tech enthusiasts who are involved in building\, maintaining\, or s caling data platforms. These individuals are likely to have a foundational understanding of data infrastructure and are looking to learn from real-w orld implementations and challenges.\n\n\nThe session aims to address the common challenges organisations face in building and scaling a data platfo rm. These include:\n\n* Handling rapid growth in data volume and complexit y\n* Choosing the right technologies and architectures\n* Integrating adva nced analytics and machine learning\n\nBy sharing Dream11’s journey\, th e talk will provide insights and practical solutions to these challenges.\ n\nScope of the session\nThe scope will cover the complete journey of Drea m11’s data platform from its inception to its current state. This includ es:\n\nInitial setup and foundational tools\nChallenges and solutions duri ng scaling\nThe current advanced data ecosystem\n\nHow will participants b enefit from your session?\nParticipants will benefit by gaining:\n\nA clea r understanding of the practical steps involved in building and scaling a data platform\nInsights into the challenges faced and how they were addres sed\nKnowledge of the latest technologies and best practices in the data d omain\nInspiration and actionable takeaways to apply to their own data pla tforms\n\n\n### Format:\nGiven the technical nature of the audience and th e objectives:\n\nSlides to illustrate key points and architectural diagram s\nReal-world examples and case studies from Dream11\nInteractive Q&A sess ion to address specific queries from the audience\nThe format will ensure that the session is informative\, engaging\, and directly relevant to the audience's interests and needs. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240719T112958Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/from-foundation-to-the -future-the-evolution-of-dream11s-data-platform-VMe4LzYi8T4fkaCjBw9T1i BEGIN:VALARM ACTION:display DESCRIPTION:From foundation to the future: evolution of Dream11's data pla tform in Data engineering and infrastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:RAGs to riches: transforming data retrieval with vector search DTSTART:20240713T060500Z DTEND:20240713T065000Z DTSTAMP:20260419T104847Z UID:session/XHpD2hcVUNVRenDty1odie@hasgeek.com SEQUENCE:7 CATEGORIES:30 mins talk CREATED:20240625T103306Z DESCRIPTION:With the recent technological advancement in LLM's\, embedding generation and RAGS\, there is immense interest in using this technology to solve problems like Semantic Search\, Chat Bots\, Code Graph\, Knowlegd e graph. \n\nThis has led to the development of Data Management systems wh ich can deal with large scale generation\, storage and search capability f or these vector embeddings.\n\nThis talk is targeted towards technical aud ience interested in details of Vector Search state of the art algorithms a nd challenges associated with building data management system for it.\n\n# # Outline\nThe talk will start with motivation about why RAGS compare it w ith traditional Search engine. Followed by introduction to basic mathemati cal constructs and popular algorithms. Talk will then delve into the state of art Vector Databases and then segue into the challenges associated wit h building those systems and open problems. Talk will touch upon the integ ration challenges related to building end to end application in this space .\n\nThe talk will conclude with a peek into the possible future.\n\n\n## Impact\nI work on the effort to build Native Vector Database support in Mi crosoft SQL Server. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T100221Z LOCATION:Data engineering and infrastructure BOF area (Ground floor foyer) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/vector-search-and-data bases-XHpD2hcVUNVRenDty1odie BEGIN:VALARM ACTION:display DESCRIPTION:RAGs to riches: transforming data retrieval with vector search in Data engineering and infrastructure BOF area (Ground floor foyer) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Food for thought: scaling delight at Zomato using GenAI DTSTART:20240713T060500Z DTEND:20240713T064000Z DTSTAMP:20260419T104847Z UID:session/5RTE6hDH6FJesgNVxDMVnf@hasgeek.com SEQUENCE:10 CATEGORIES:30 mins talk CREATED:20240625T112858Z DESCRIPTION:### Introduction\n\nIn today's rapidly evolving digital landsc ape\, Generative AI is playing a pivotal role in transforming how our busi nesses interact with the customers.\n\nGenAI is being leveraged to solve a variety of use-cases innovatively at Zomato and Blinkit - Recipe Rover\, Zomato AI\, People Co-pilot\, Logs Co-pilot\, and Customer Delight (ZIA).\ n\nThis talk will focus on the journey of Gen AI in scaling Customer Delig ht at Zomato.\n\n### Target Audience\nThe primary audience for this talk i ncludes anyone interested in Gen AI applications at scale - AI researchers \, developers\, engineers\, product and business leaders.\n\n### Outline F or Customer Delight Solution Deep Dive\n\n- Problems with old flow\n - Fe els transactional and pre-defined\n - Unable to decipher intent and actua l issue faced by the customer\n- Requirements of a solution\n - Feel conv ersational and generate contextual replies\n - Understand the intent and issue and provide resolutions accordingly\n - Escalate to a human agent w hen unable to provide resolution\n- Prime metrics to track success of the solution\n - CSAT scores\n - Escalation %\n - Cost and Latency\n- AI So lution Design\n - Explanation and reasoning behind chosen architecture\n - Marrying linguistic variables with code - policies\n - Some other opti misations\n - Evaluation pipeline for staying on top\n- Next Steps\n - O ptimizing for control\, cost\, latency\n - Personalisation for customer l ove\n - Automating post order delivery delight\n - Expand to Blinkit\n\n ### Outline For System Design to Enable AI Customer Delight at Scale\n\n- Agent Orchestrator Service\n- Internal AI Interfaces Demo\n- Internal LLM Evaluation Toolkit\n\n### Key Takeaways\n\n- Building and scaling GenAI sy stems for real-world applications (Mother's day performance of the bot)\n- Transforming traditional customer support from transactional to conversat ional\n- Integrating linguistic models with business policies to implement guardrails in AI systems\n- Evaluation pipeline for a constant feedback l oop GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071212Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/scaling-customer-delig ht-at-zomato-using-ai-5RTE6hDH6FJesgNVxDMVnf BEGIN:VALARM ACTION:display DESCRIPTION:Food for thought: scaling delight at Zomato using GenAI in LLM s and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T063500Z DTEND:20240713T064500Z DTSTAMP:20260419T104847Z UID:session/V6eYfFYHFFz5ceMYJ5wpJW@hasgeek.com SEQUENCE:2 CREATED:20240701T095408Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T095413Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in Workshop\; BOFs for LLMs and GenAI track (Librar y on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T064000Z DTEND:20240713T065000Z DTSTAMP:20260419T104847Z UID:session/WeJL6LuAdrEuWaesZ3EBpR@hasgeek.com SEQUENCE:3 CREATED:20240625T100931Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054318Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in Data engineering and infrastructure track (Semin ar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T064000Z DTEND:20240713T065000Z DTSTAMP:20260419T104847Z UID:session/4hfCHBtfTUFoQ1egFHtsaR@hasgeek.com SEQUENCE:3 CREATED:20240625T112941Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054320Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in LLMs and GenAI track (Auditorium on the ground f loor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Birds of Feather (BOF): Deploying AI in key sectors: robust risk m itigation strategies DTSTART:20240713T064500Z DTEND:20240713T073000Z DTSTAMP:20260419T104847Z UID:session/9Fy1vPtfk6hzdcn9roBHW5@hasgeek.com SEQUENCE:11 CATEGORIES:Birds of Feather (BOF) session CREATED:20240625T112751Z DESCRIPTION:## Abstract:\nAs AI continues to revolutionize various sectors \, it brings both unprecedented opportunities and significant risks. In In dia\, sectors such as Agritech\, Fintech\, Edtech\, public services\, and Healthtech are rapidly adopting AI technologies. However\, the lack of rob ust risk mitigation strategies can lead to unintended consequences\, inclu ding data breaches\, algorithmic biases\, and systemic vulnerabilities. \n \nOver the past year\, we have spoken to practitioners and researchers and collated research on the unique risks in each sector and spicific mitigat ion strategies. This talk outlines these risks and mitigation strategies a s well as best practices for regulatory compliance.\n\n## Outline:\nThe se ssion will be in an open discussion with 4-5 panelists. The structure will broadly cover the following:\n- **Introduction**: Overview of AI adoption in Agritech\, Fintech\, Edtech\, public services\, and Healthtech in Indi a.\n- **Risks in AI Deployment**: Identification of key risks associated w ith AI in each sector.\n- **Mitigation Strategies**: Presentation of effec tive risk mitigation strategies specific to each sector\, with an emphasis on practical implementation.\n- **Policy and Compliance**: Discussion on the role of policymakers and compliance teams in implementing these strate gies.\n- **Best Practices**: Sharing best practices and guidelines for dev elopers\, researchers\, and startup founders to mitigate AI risks.\n\n## K ey participants:\nDevelopers\, researchers\, policymakers and analysts\, s tartup founders\, risk and compliance teams.\n\n## Impact:\n- Gain a compr ehensive understanding of the risks associated with AI deployment in criti cal sectors in India.\n- Learn actionable risk mitigation strategies that can be implemented in real-world scenarios.\n- Obtain insights into the po licy landscape and how to navigate compliance requirements effectively.\n- Equip participants with knowledge to proactively address potential risks\ , ensuring sustainable and ethical AI practices. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240706T114910Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/ai-and-risk-mitigation -strategies-in-key-indian-sectors-9Fy1vPtfk6hzdcn9roBHW5 BEGIN:VALARM ACTION:display DESCRIPTION:Birds of Feather (BOF): Deploying AI in key sectors: robust ri sk mitigation strategies in Workshop\; BOFs for LLMs and GenAI track (Libr ary on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Establishing Causality in Complex Mental Health Issues using AI DTSTART:20240713T065000Z DTEND:20240713T073000Z DTSTAMP:20260419T104847Z UID:session/Bag3rAHV7sKyp1XAskTw5x@hasgeek.com SEQUENCE:6 CATEGORIES:30 mins talk CREATED:20240625T113001Z DESCRIPTION:This talk explores the forefront of artificial intelligence (A I) in establishing causality in mental health. By leveraging Graph Neural Networks (GNNs) and Spatio-Temporal Graph Neural Networks (STGNNs)\, we ai m to uncover causal relationships in complex mental health causal effects. The session will cover fundamental concepts of causality\, the transition from traditional GNNs to STGNNs\, and the creation of Clinical and User K nowledge Graphs (KGs). We will also delve into the novel GNN-RAG approach for enhancing reasoning in large language models and discuss the ethical\, privacy\, and regulatory challenges. Collaboration with healthcare profes sionals and stakeholders will be emphasized to ensure practical and ethica l implementation.\n\n## Outline\nIntroduction to Causality in Mental Healt h\n\n- Importance of causality in understanding mental health.\n-- Traditi onal vs. AI-driven approaches.\n-- Fundamentals of Causality\n\n- Overview of causality concepts.\n-- Methods for causal inference: correlation vs. causation.\n-- Graph Neural Networks (GNNs) in Causal Inference\n\n- Intro duction to GNNs and their applications.\n-- How GNNs help in causal discov ery and inference.\n-- Case studies in mental health using GNNs.\n-- Time + Directionality = Causality\n\n- The role of temporal and directional dat a in establishing causality.\n-- Techniques to model time and directionali ty in mental health data.\n-- From GNNs to Spatio-Temporal Graph Neural Ne tworks (STGNNs)\n\n- Evolution from GNNs to STGNNs.\n-- How STGNNs enhance causal analysis by incorporating spatial and temporal dimensions.\n-- App lications in mental health scenarios.\n-- Creating Clinical and User Knowl edge Graphs (KGs)\n\n- Building and utilizing KGs to represent clinical an d user data.\n-- Integration of KGs with GNNs and STGNNs for richer causal insights.\n-- GNN-RAG: Graph Neural Retrieval for Large Language Model Re asoning\n\n- Overview of GNN-RAG methodology.\n-- Enhancing reasoning capa bilities of large language models using GNN-RAG.\n-- Implications for ment al health analysis.\n-- Privacy\, Ethics\, and Regulatory Considerations\n \n- Key privacy concerns with AI in mental health.\n-- Ethical issues and the need for transparent AI models.\n-- Navigating regulatory landscapes a nd ensuring compliance.\n-- Collaboration with Healthcare Professionals & Other Stakeholders\n\n- The importance of interdisciplinary collaboration. \n-- Strategies for effective communication and implementation of AI in cl inical settings. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240719T111827Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/establishing-causality -using-ai-in-mental-health-Bag3rAHV7sKyp1XAskTw5x BEGIN:VALARM ACTION:display DESCRIPTION:Establishing Causality in Complex Mental Health Issues using A I in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Sponsored talk: The winter of our discontent: content moderation s ystems at scale DTSTART:20240713T065000Z DTEND:20240713T073000Z DTSTAMP:20260419T104847Z UID:session/FkwPr9aVMT99rmm3kXm3hC@hasgeek.com SEQUENCE:18 CATEGORIES:30 mins talk CREATED:20240625T100621Z DESCRIPTION:We heavily rely on the Web for meeting our information needs t oday. Examples include Wikipedia\, Twitter\, Instagram\, Youtube\, Google Maps etc. All of these are platforms where millions of users post billions of pieces of content every day on a wide range of topics. The content is consumed by hundreds of millions of users. While a rich source of informat ion\, these platforms are also easy targets for abuse and harm\, both inte ntional as well as un-intentional. Intentional harm includes the use of th ese platforms for fraud\, misinformation\, trolling\, hate and other forms of vandalism. Unintentional harm includes factually incorrect information \, stale information\, information bias and more.\nThe underlying platform providers have a huge responsibility in terms of ensuring that users are provided a safe\, delightful\, useful and transparent experience for the i nformation that is presented to them. However\, this is a very difficult p roblem to solve in the real world. There are a large number of very hard c hallenges to deal with. \nFor example:\nHow do we deal with highly motivat ed bad actors who are technically savvy\, have financial means and put con tinuous efforts to identify vulnerabilities and expose them\, adapting at a very fast pace?\nHow do we differentiate between facts and opinions and determine the actual “ground truth label” and do it at scale\, with th orough representation for all kinds of patterns and do it fast before the patterns change so that our models have a good exposure to the underlying problem\nHow do we deal with signal sparsity? What if we simply don’t ha ve any useful signals to train the classifier on?\nWhat do we do if the to lerance for errors and mistakes is very low and yet\, we don’t have an e ffective classifier to solve the problem?\nHow do we solve this for a larg e number of semantically different types of documents?\nHow do we build th e system to deal with billions of documents and build an efficient and rel iable system?\n\n\nIn this short talk\, we will elaborate on this super cr itical and pressing challenge faced by the tech industry and give you insi ghts into how the above challenges and many more are solved in the GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071310Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/content-moderation-sys tems-at-scale-FkwPr9aVMT99rmm3kXm3hC BEGIN:VALARM ACTION:display DESCRIPTION:Sponsored talk: The winter of our discontent: content moderati on systems at scale in Data engineering and infrastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Lunch break DTSTART:20240713T070000Z DTEND:20240713T080000Z DTSTAMP:20260419T104847Z UID:session/Xh83oFhnTp2aGGKH2Cd8eF@hasgeek.com SEQUENCE:5 CREATED:20240701T100556Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T055237Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Lunch break in Board room (second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Lunch DTSTART:20240713T073000Z DTEND:20240713T083000Z DTSTAMP:20260419T104847Z UID:session/2MjdKWZPYWg6cxcn1QuQKy@hasgeek.com SEQUENCE:8 CREATED:20240625T101024Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240704T102600Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Lunch in Data engineering and infrastructure track (Seminar ha lls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Lunch DTSTART:20240713T073000Z DTEND:20240713T083000Z DTSTAMP:20260419T104847Z UID:session/CLZw7yFB19SztCQkvxZM9z@hasgeek.com SEQUENCE:11 CREATED:20240625T112804Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T095425Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Lunch in Workshop\; BOFs for LLMs and GenAI track (Library on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Lunch DTSTART:20240713T073000Z DTEND:20240713T083000Z DTSTAMP:20260419T104847Z UID:session/Hnb9v4xTNa43P9LZdDXdcm@hasgeek.com SEQUENCE:6 CREATED:20240625T113111Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T054413Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Lunch in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Building Data Products to Empower Entrepreneurial Execution DTSTART:20240713T080000Z DTEND:20240713T093000Z DTSTAMP:20260419T104847Z UID:session/YG9ZcvjYCQwyF78QsMwpHs@hasgeek.com SEQUENCE:7 CATEGORIES:Workshop CREATED:20240625T114017Z DESCRIPTION:Build a Data Product - A Roundtable Discussion\n\n## Outline\n The primary pain point I am addressing is the gap between having a technic al skill set and being able to transform those skills into a viable data p roduct. Many technologists excel in coding and technical configurations bu t struggle with:\n* Identifying market needs and opportunities where data can add value.\n* Understanding the various moving parts required to build a product that can be accessed by users on the Internet.\n* Managing the full lifecycle of a data product from conception through to deployment and scaling.\n\nThe rountable session seeks to create a forum where the above items can be discussed from a variety of perspectives.\n\n## Impact\nThe impact I plan to drive for participants of this round table are the follow ing\;\n* Practical Knowledge: Learn how to identify viable opportunities f or data products in various industries.\n* Technical Guidance: Insights on using familiar technologies to build and deploy data products effectively .\n* Networking Opportunities: Engage with like-minded individuals who are also interested in building data products\, which can lead to potential c ollaborations.\n* Inspiration and Motivation: Stories and examples from re al-world projects that demonstrate the potential and impact of well-execut ed data products.\n\nThis session will help bridge the gap between technic al ability and entrepreneurial execution\, empowering participants to not only build data products but also to understand the ecosystem in which the se products must thrive.\n\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T055238Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/build-a-data-product-a -roundtable-discussion-YG9ZcvjYCQwyF78QsMwpHs BEGIN:VALARM ACTION:display DESCRIPTION:Building Data Products to Empower Entrepreneurial Execution in Board room (second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Imagining the Future of AI in India DTSTART:20240713T081000Z DTEND:20240713T091500Z DTSTAMP:20260419T104847Z UID:session/WQZPenUeprasp1Rjfb2HkK@hasgeek.com SEQUENCE:3 CATEGORIES:Birds of Feather (BOF) session CREATED:20240712T121528Z DESCRIPTION:The rush for building AI is taking over every business organiz ation to adapt to a new era of automation. The push for AI is going to tak e over the economy and society at large with applications of AI in every s ector. This brings many important questions of production of AI\, from lar ge scale data centers to data sets required to train these complex mathema tical models. In this context\, how do we imagine our future and how shoul d the community build futures that are just.\n\nThis birds of feather sess ion will look into critical questions of AI beyond the standard questions of ethics and fairness in AI. This session is aimed to bring out critical discussions on what AI futures we want and how one should go about it. Wha t kind of philosophy should the community evolve to tackle new challenges of AI? Would AI really bring a terminator era or will it be more like the creator\, what is the role of science fiction in our AI imaginations? Woul d a FOSS model of AI be enough in a world where large computers will deter mine futures? What about the role of privacy\, censorship and digital righ ts in AI? Should AI be regulated? If so\, at what point?\n\nThese are some important questions that need to be addressed by society as it adopts AI into every aspect of our lives. Where do we start and how do we go about i t?\n\n## Readings:\n\nThe Future of Artificial Intelligence and Cybernetic s https://www.technologyreview.com/2016/11/10/156141/the-future-of-artific ial-intelligence-and-cybernetics/\n\nAI Should Augment Human Intelligence\ , Not Replace It\nhttps://hbr.org/2021/03/ai-should-augment-human-intellig ence-not-replace-it\n\nAI is an energy hog. This is what it means for clim ate change. https://www.technologyreview.com/2024/05/23/1092777/ai-is-an-e nergy-hog-this-is-what-it-means-for-climate-change/\n\nShould We Stop Deve loping AI For The Good Of Humanity? https://www.forbes.com/sites/bernardma rr/2023/05/03/should-we-stop-developing-ai-for-the-good-of-humanity/\n\nTh e role of philosophy in mathematical modeling as understood by scientists at TCS https://archive.org/details/complex-problems-and-qualitative-modell ing\n\nScience Fiction Imaginations of Information https://www.wired.com/1 995/11/krauss/ GEO:12.9666826;77.6352903 LAST-MODIFIED:20240712T121550Z LOCATION:Terrace (third floor) - Bangalore International Centre\nBangalore \nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/imagining-the-future-o f-ai-in-india-WQZPenUeprasp1Rjfb2HkK BEGIN:VALARM ACTION:display DESCRIPTION:Imagining the Future of AI in India in Terrace (third floor) i n 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Starting from scratch: lessons learnt from building high performan ce parallel distributed compute platforms DTSTART:20240713T083000Z DTEND:20240713T090500Z DTSTAMP:20260419T104847Z UID:session/RTmADhCdBp5wXd9JicVZfa@hasgeek.com SEQUENCE:12 CATEGORIES:30 mins talk CREATED:20240625T101005Z DESCRIPTION:## Introduction\nPlatform engineering and data architecture te ams are increasingly adopting object-store backed data lakehouses as their central\, unified platform for workloads across Analytics as well as AI.\ n\nWith the scale of such data lakehouses ranging from the 10s of TBs to t he 100s of PBs\, distributed compute engines like Spark\, Trino / Presto\ , Flink\, etc. are essential for workloads across: \n- Data ingestion \n- Data transformation / pre-processing \n- Data Querying / inference \n\nThi s talk covers the common challenges data platform teams encouter with popu lar distributed compute engines at scale. \n\nWe then outline our approach to building a new class of hyper-efficient compute engine from scratch. W e also outline how this new approach provides substantial advantges in a c lass of technically challenging workloads that combine one or more of:\n1. High concurrency\n2. Query Complexity\n3. High Data Volumes\n4. Stringent latency requirements\n\nThe talk will have a mix of presentation (slides) \, benchmarking\, live demos\, and audience Q&A.\n\n## Target Audience\nEn gineers\, researchers and data architects with an interest in:\n- Massivel y parallel distributed compute platforms\n- The internals of existing and emerging compute engines\n- Composable open data platforms with an emphasi s on object-store based data lakes and lakehouses with open table formats like Delta lake\, Iceberg\, and Hudi\n\n## Outline\nWith a query's lifecyc le as the frame of reference\, we start with examining the strengths and w eaknesses of the present engines.\n\nWhile most distributed compute engine s are available as Open Source (OSS) as well Commercial Open Source Softwa re\, all of them share commonalities on the following areas:\nA - Monolith ic\, stateful and "VM-centric" Architectures\nB - Centralized and static a pproach to distributed processing and execution\n\nWe then present how a c lean-slate approach helped us build a system that overcomes the key limita tions through the use of:\nA - Disaggregated\, stateless\, and "kubernetes -native" Architecture\nB - Decentralized and dynamic approach to distribut ed processing and execution\n\n## Takeaways\, Impact \nWe will present fin dings from real-world workloads around how this new approach drives benefi ts across evaluation criteria that matter to platform engineering teams:\n \n1 - A materially superior Price-Performance curve\n2 - Eliminating syste m-wide Single Points of Failure (SPOF)\n3 - Maintaining Deterministic tail latencies (p99) even under heavy loads and massive variability\n4 - Effic ient cluster utilization even when faced with data skew\, variable task co mpletion\, etc. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T082216Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/a-new-approach-to-buil ding-high-performance-lakehouse-compute-engines-for-open-table-formats-lik e-delta-lake-apache-iceberg-and-apache-hudi-RTmADhCdBp5wXd9JicVZfa BEGIN:VALARM ACTION:display DESCRIPTION:Starting from scratch: lessons learnt from building high perfo rmance parallel distributed compute platforms in Data engineering and infr astructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Leveraging Llama.lisp for constructing compiler frameworks DTSTART:20240713T083000Z DTEND:20240713T090500Z DTSTAMP:20260419T104847Z UID:session/2Vb1QxYRkzgWSHD2qHv3qt@hasgeek.com SEQUENCE:7 CATEGORIES:30 mins talk CREATED:20240625T113704Z DESCRIPTION:\n## Abstract:\n\n\nCompilers are workhorses of performance be hind all AI algorithms. Making algorithms work effectively on GPUs is espe cially hard - called kernel programming. Compiler ecosystem around GPUs is especially messed up. Compilers are supposed to allow for performance por tability of different hardwares but this is usually not the case. See belo w infographic for current state of AI compilers. \n\n ![https://twitter. com/AartBik/status/1782216468006138312](https://pbs.twimg.com/media/GLu0Qs waUAA-wVY?format=jpg&name=large)\n\nI am designing a compiler framework ca lled llama.lisp. As you can tell from the name\, the framework is highly i nspired by Lisp. Limitations of LLVM are examined critically to improve th e framework. A multi layered approach is adopted to tame the complexity of writing such an compiler. No single language is chosen as implementation language to allow for the right tools can be used for right problem. A mes sage oriented architecture is used for foreign function interfacing instea d of traditional C based API.\n\nIn this talk\, I will go through design c hoices of this compiler framework.\n\n## Additional Information:\n\n**Audi ence for the session**\n\nFolks who are interested in GPU programming\, sy stems programming. Their roles might currently be called as Data engineer\ , AI engineer or systems engineer. Experience with high performance comput ing or array programming would make you appreciate the talk better.\n\n\n* *Scope of the session**\n\nWe'll go through high level design decisions wi thout digging too much into details. We might look at some scary assembly code but don't expect you to read it.\n\n\n**Takeaways from the sessions** \n\nBy the end of this talk\, you will have a good idea of computer archit ectures of AI and how programming languages are implemented.\n\n\n**Format of the session**\n\n30 mins talk with 10/15 mins QnA GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T065513Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/llama-lisp-design-of-a n-ai-first-compiler-framework-2Vb1QxYRkzgWSHD2qHv3qt BEGIN:VALARM ACTION:display DESCRIPTION:Leveraging Llama.lisp for constructing compiler frameworks in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Fetch Data from Data Lakes by Building a Low-Code Data Assistant f or Businesses using Code LLMs DTSTART:20240713T083000Z DTEND:20240713T093000Z DTSTAMP:20260419T104847Z UID:session/7ck37VHbTJwWWYYu7ggCHW@hasgeek.com SEQUENCE:6 CATEGORIES:Birds of Feather (BOF) session CREATED:20240625T113929Z DESCRIPTION:This BoF session talks about building low-code data assistant for business using code LLMs. Generative AI (LLMs) for codes had become ve ry popular and powerful tool for developers to leverage with rise of enter prise solution like GitHub Copilot\, AWS Code Whisperers\, Google Duet etc along with numerous open source code assist models for generating codes f or hundreds of programming languages including python\, java\, SQL etc. Th ese models can take english comments or instructions to write code for dev elopers. \n\nApart for improving productivity of developers\, these models can also be used to create impactful NL2SQL solution in data space.\n* Al lows business users (non-tech) to use natural language to answer day to da y queries.\n* Speed up data analysis for insights & report generations.\n\ nCEO of Snowflake Inc. (cloud-based data company)\, Sridhar Ramaswamy in p ress briefing said\, “Our dream here is\, within a year\, to have an API that our customers can use so that business users can directly talk to da ta\,”. Natural language querying can enables various stakeholders — in cluding executives\, employees\, customers\, prospects\, and partners — to pose questions about data in natural language and receive relevant resp onses.\n\n### Who is the Audience ?\n* *Data enginners\, Data scientists o r researchers* who are closer to enterprise data.\n* *Business analysts or Business consumers* who require data querying and analysis on day to day basis.\n* *Data leaders* as an initiative to democratize data for enterpri ses.\n\nThe talk will align to any professional working on LLMs & Gen AI u secases.\n\n### What problem are you trying to solve (for the audience)?\n * Many business users (non-tech) often struggle to fetch data from data la kes and systems which leads to higher effort from data teams to produce th e same on regular basis. How can we automate the same ?\n* Many business a nd financial analyst perform ad-hoc analysis to answer business queries on the fly on downloaded Excel\, CSVs data. How can we automate it ?\n* Data governance teams in enterprise validates data migration process with many test cases to ensure data quality. How can we reduce extensive effort in writing data queries ?\n* Data governance teams to create metadata in term s of sample SQL queries for Data Catalog.\n* The talk helps data scientist s & data teams on going about the implementing such solution with right to ols & techniques.\n\n## Scope\n1. ***Enroads to Enterprise***: Introductio n to an enterprise problem\, potential use-cases and application of NL-2-S QL.\n * Adhoc Data Queries: Automation\n * Validation in Data Migr ation\n * Enriching Data Catalog\n \n2. ***State of Code LLMs***: \n * Closed vs Open models.\n * Generalist vs Specialist models\n3. ***Practical challenges***: Potential challneges in scaling and strategie s in rescue.\n * Multi-Table scenario\, How to find right tables in Sno wflake ? \n * Few shot - Retrieval examples.\n * Metadata pruning.\ n * Hallucination correction (SQL Glot)\n * Instructions in Prompt\n * Fine-Tuning\n \n5. ***Building Data Assistants***: Implementa tion components & process flow.\n \n### Benefit & Key takeaways for Part icipants\n* A potential idea to democratize data to Business or Enterprise .\n* Practical details on solutioning the low code/No code data assistant capability to make data closer to business.\n* Current state of Code gener ative Large Language Models.\n* Indepth insights in effort and challenges in creating NL2SQL solution. \n* Techniques and Tools to implement such s olution and overcome challenges.\n\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240705T110631Z LOCATION:Data engineering and infrastructure BOF area (Ground floor foyer) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/unlock-data-with-nl2sq l-building-low-code-data-assistant-for-business-using-code-llms-7ck37VHbTJ wWWYYu7ggCHW BEGIN:VALARM ACTION:display DESCRIPTION:Fetch Data from Data Lakes by Building a Low-Code Data Assista nt for Businesses using Code LLMs in Data engineering and infrastructure B OF area (Ground floor foyer) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Opening Up Open Source with Indic Language LLMs DTSTART:20240713T083000Z DTEND:20240713T090500Z DTSTAMP:20260419T104847Z UID:session/KuWShFqPbSKGu4axyGFYLY@hasgeek.com SEQUENCE:10 CATEGORIES:30 mins talk CREATED:20240625T113215Z DESCRIPTION:Introduction\n\nLike steam engine\, Electricity and Internet b ecame integral to the first\, second and third Industrial revolutions\, A I is going to be adapted sooner than later in all production and business processes. It is very rapidly going to change the way people conduct their businesses and how the production processes are executed. \nWhile the Fre e/open and the proprietary/closed nature of the source code in the third i ndustrial has been resolved with Free/Open model taking the centre stage a nd fostered the faster penetration of IT and its adaption. \nIn world of A I\, source code is only one of the four components among the algorithm\, t he source code (or code)\, the data\, and the model. Non exclusivist aspec ts of algorithm\, Openness of algorithmic implementation ie verifiability of the model to construe to the algorithm\, diversity of data sets become crucial.\nData is the new oil apparently. And just as with oil\, we are se eing lots of techniques being employed to collect this data. Are these dat asets being collected with permissions from the original creators. Does th e voice of the creator matter. As we see in the current fiasco of the Scar lett Johansson voice in OpenAI demo\, rights matter. Training this data n eeds humungous amounts of compute\, by some estimates\, the power needed t o train these models is as much as that needed to run a city.\nThis talk w ill focus on the tryst with an alternative path of making Datasets Collabo ratively\, building models responsibly and ensuring that the licenses of t he datasets protect the datasets in the interest of the commons.\n\nTarget Audience\nThe primary audience for this talk includes anyone interested i n AI who want to solve problems in collaboration - AI researchers\, develo pers\, engineers\, product and business leaders.\n\n\n \n## Outline\n Dat a the new Oil and with the Digital Monopolies or Governments\n ◦ Data Harvesting\n ◦ Privacy issues\n • The alternative\n ◦ Datathons\n ▪ Swecha Chandamama Kathalu\n ▪ Voice samples for ASR\n ▪ 100000 internships\n ◦ The Compute\n ◦ The Model\n • Prime issue of licensing\n ◦ Source Code\n ◦ Model\n ◦ Datasets\n • Licensing of Data Sets\n ◦ The issue with the current licenses\n ◦ The alternative license\n\n## Key Takeaways\n\n A working exam ple of collecting data responsibly and building models in the open\n A direction to think in terms of licenses applicable for AI datasets and AI models\n How to build an AI for the people\, by the people.\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T070802Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/ai-by-the-people-for-t he-people-KuWShFqPbSKGu4axyGFYLY BEGIN:VALARM ACTION:display DESCRIPTION:Opening Up Open Source with Indic Language LLMs in Workshop\; BOFs for LLMs and GenAI track (Library on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T090500Z DTEND:20240713T091500Z DTSTAMP:20260419T104847Z UID:session/M2WDQ2gCAMcPaCezcnXavD@hasgeek.com SEQUENCE:2 CREATED:20240625T113729Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T055155Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in LLMs and GenAI track (Auditorium on the ground f loor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T090500Z DTEND:20240713T091500Z DTSTAMP:20260419T104847Z UID:session/PCzgCVuYcZGM3KNe1zaF5h@hasgeek.com SEQUENCE:4 CREATED:20240625T123513Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240625T123546Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in Workshop\; BOFs for LLMs and GenAI track (Librar y on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T090500Z DTEND:20240713T091500Z DTSTAMP:20260419T104847Z UID:session/HKyJQ3M6cs22NRqTgk6QV6@hasgeek.com SEQUENCE:3 CREATED:20240625T101938Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240704T102611Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in Data engineering and infrastructure track (Semin ar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Unlocking the power of Real Time Feature Stores DTSTART:20240713T091500Z DTEND:20240713T095000Z DTSTAMP:20260419T104847Z UID:session/Lfp1DcZFWPZbfm6GMnAdyi@hasgeek.com SEQUENCE:4 CATEGORIES:15 mins talk CREATED:20240706T120053Z DESCRIPTION:In today's data-centric world\, businesses rely on personaliza tion now more than ever. Whether it's personalizing user experiences\, opt imizing operations\, or predicting market trends\, data plays a pivotal ro le. To harness the full potential of data\, organizations are turning to r eal-time feature stores. In this talk\, we'll explore what real-time featu re stores are\, why they matter\, and how we at Zupee have leveraged real- time feature stores for various use cases.\n\nZupee is a rapidly growing r eal money gaming platform that caters to a growing user base of more than 100 million registered users. We have been building data powered models to enhance the overall user experience and improve the integrity solutions o n our app. These include multiple ML models to improve user recommendation s\, system integrity solutions etc. The full potential of these models can be leveraged when deployed on real-time data. Serving features on a relia ble\, real-time and centralized basis would be necessary for these to work . This calls for a Real Time Feature store to be built.\n## Outline\n- The audience for this talk would be Data Engineers\, Data Scientists or ML En gineers who are building platforms to provide personalized experiences in real time\n- To Build a reliable and scalable real time data infrastructur e that powers ML use cases is incredibly challenging to build and this tal k would be helpful to folks who are thinking of building such user experie nces\n- Providing guidance on high level design and talking though the cha llenges in execution so that folks can avoid the common pitfalls while bui lding real time feature stores\n\n## Impact\nBy leveraging the real-time f eature store\, we are able to activate multiple ML use cases such as fraud detection\, player protection\, personalized experiences\, recommender en gines etc. on a real-time basis and also serve these centralized features for other use cases such as:\n1. Enhanced Decision-Making\n2. Efficient Mo del Training\n3. Ensuring Consistency\, Reliability and Reproducibility\n4 . Enabling Governance GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T082526Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/unlocking-the-power-of -real-time-feature-stores-Lfp1DcZFWPZbfm6GMnAdyi BEGIN:VALARM ACTION:display DESCRIPTION:Unlocking the power of Real Time Feature Stores in Data engine ering and infrastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Birds of Feather (BOF): Licenses to Level the Playing Field in Gen AI DTSTART:20240713T091500Z DTEND:20240713T101500Z DTSTAMP:20260419T104847Z UID:session/UvCzuFSRuGiXyC6t1rSSn3@hasgeek.com SEQUENCE:8 CATEGORIES:Birds of Feather (BOF) session CREATED:20240625T113356Z DESCRIPTION:## Introduction\nIn this rapidly evolving digital era\, data a cts as the fuel powering the relentless growth of artificial intelligence. As we stand on the brink of technological revolutions\, it becomes crucia l to understand not just how data drives AI\, but also the ethical and leg al frameworks that must evolve with it. We should try to look at licensing as a tool to make sure that we can level the playing field where currentl y because of the access to data and compute the incumbents are reaping mos t of the benefits of AI.\n\nIn this talk we will walk through our journey of trying to find the right license for the crowd sourced data collected f or the Telugu ASR system and the model built using the data.\n## Target Au dience\nThe primary audience for this talk involves Data Scientists and AI Researchers\, Legal Professionals with a Focus on Technoloy\, Open Source and Community Contributors\, Policy Makers and Regulators.\nAcademics and Students in Technology and Law Fields\, Tech Entrepreneurs and Start-up F ounders can also benefit from this talk.\n## Outline\n- Introduction to th e Project\n - Overview of the Telugu ASR system\n - Importance of cr owdsourced data in ASR technologies\n\n- Challenges with Licensing Crowdso urced Data\n - Legal complexities of using crowdsourced data\n - Eth ical considerations in data collection and usage\n\n- Requirements for an Effective License\n - Compliance with data protection regulations (e.g.\, GDPR\, CCPA)\n - Flexibility to accommodate contributions from a diverse crowd\n - Clarity on data usage rights and restrictions\n\n- Journey to Finding the Right License\n - Evaluation of existing licenses (e.g.\, C reative Commons\, MIT\, proprietary licenses)\n - Customizing license e lements to suit specific needs of datasets for AI which will take care of new terms like fine tuning\, model weights etc\n - Engagement with lega l experts and the community\n\n\n- Next Steps and Future Directions\n - Open house for consultations\n - Finalizing the license and release\n- Q &A\n - Open floor for questions and further discussion\n\n## Impact\nIntr oducing a specialized license for crowdsourced data\, akin to the impact o f the GPL for open-source software\, could fundamentally transform how dat a is utilized in technological innovations. It would promote a collaborati ve environment where data can be freely shared and enhanced\, while ensuri ng compliance with ethical standards and data protection laws. Such a lice nse would encourage broader participation and innovation\, reduce legal ba rriers\, and ensure the sustainability of data resources. It might also he lp level the playing field by making sure the benefits dont accrue to only the mega corporations in AI. By clarifying usage rights and responsibilit ies\, this new licensing framework could set industry standards for data h andling\, leading to more responsible and impactful technological advancem ents across various sectors. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T110614Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/need-for-new-licenses- in-this-age-of-generative-ai-UvCzuFSRuGiXyC6t1rSSn3 BEGIN:VALARM ACTION:display DESCRIPTION:Birds of Feather (BOF): Licenses to Level the Playing Field in GenAI in Workshop\; BOFs for LLMs and GenAI track (Library on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Making JIRA more Agile with AI DTSTART:20240713T091500Z DTEND:20240713T095000Z DTSTAMP:20260419T104847Z UID:session/94Nx3NDYRQam9GZd9XyCck@hasgeek.com SEQUENCE:6 CATEGORIES:30 mins talk CREATED:20240625T113753Z DESCRIPTION:\n## Introduction\n\nAtlassian's Jira Service Management (JSM) has consistently strived to empower customers by delivering top-notch ass istance to those in need. A primary objective has been to promote self-ser vice within JSM\, allowing users to promptly access help while reducing th e workload on agents.\n\nThe introduction of Generative AI led to the deve lopment of JSM AI Answers\, enhancing JSM's search functionality to provid e conversational and precise responses. This talk will delve into the tran sition from conventional search systems to an advanced semantic search sys tem\, bolstered by large language models.\n\n## Target Audience\nThe talk is intended for a wide audience\, encompassing AI researchers\, developers \, engineers\, as well as product and business leaders\, who are keenly in terested in large-scale Gen AI applications.\n\n## Outline\n* Anatomy of JSM Help\n * Involvement of Agents and Help Seekers\n * Support tickets\n * Request Types\n * Knowledge Base\n* JSM Search Ev olution\n * Keyword-based search\n * Limitations in understa nding user queries and intent\n * Lack of personalized results ba sed on user profiles\n * AI-based search\n * RAG architectur e (Retrieval Augmented Generation) and its adaptation for this use-case\n * Help Resources - 1P (First Party powered via Confluence) and 3P (Go ogle Drive\, Sharepoint\, Dropbox\, etc.)\n* Defining success metrics \n * Number of Distinct Knowledge Base Articles\n * CSAT\n * Assistance and Resolution Rate\n* Measuring Quality\n * Evaluation framework - LLM evaluator vs non-LLM evaluation \n * Answer Qu ality \n * Content Gap Analysis \n* Improving Answer Quality - Achieving 94% Accuracy \n * Enhancing query understandin g and utilizing multiple-query variations to improve search relevancy \n * Leveraging bi-encoders and cross-encoders for document rankin g and re-ranking \n * Personalizing content based on user loca tion \n * Addressing vague queries \n * Enhancing search result consistency by eliminating user-based personalization \n * Identifying and reducing hallucinations using Chain-of-Thought and System-2-Attention \n * Recommendations for request types \n* Current performance \n * Customer usage and feedback \n * Latency improvements - Reducing from 23s to 9s \n* Next Steps \n * Seamless transition between Virtual Agent and A I Answers \n * Gap Analysis - Assisting customers in identifyi ng knowledge base system gaps \n * AI-powered Knowledge Base C reation - Aiding agents in writing/updating KB articles on content gaps \n## Key Takeaways\n* Application of RAG and LLMs on a large scale an d the architectural variances from academic standards\n* Challenges and in sights in making iterative improvements and efficient evaluation\n* Diffic ulties in objectively measuring the system's "helpfulness" to users\n* Ext ending and enhancing the feature by incorporating human feedback for conti nuous improvement. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071128Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/unified-help-in-jira-s ervice-management-using-ai-94Nx3NDYRQam9GZd9XyCck BEGIN:VALARM ACTION:display DESCRIPTION:Making JIRA more Agile with AI in LLMs and GenAI track (Audito rium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Evening break DTSTART:20240713T093000Z DTEND:20240713T100000Z DTSTAMP:20260419T104847Z UID:session/TwNyrp9kKiS15v864xou2F@hasgeek.com SEQUENCE:1 CREATED:20240701T100645Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T100649Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Evening break in Board room (second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Enterprise-ready data life cycle: powering AI and analytics at sca le DTSTART:20240713T093500Z DTEND:20240713T103500Z DTSTAMP:20260419T104847Z UID:session/SMcBc8YCPW7NFkB6fkSmaF@hasgeek.com SEQUENCE:4 CATEGORIES:Birds of Feather (BOF) session CREATED:20240705T110716Z DESCRIPTION:In this session\, we discuss _Atlassian data architecture_ to help demystify the complexities around building a real-world scalable Delt a Lakehouse meeting data governance and compliance requirements and how we enabled various teams to iterate fast for their data-driven initiatives.\ n\nToday companies are increasingly looking to make their data ever so acc essible to fuel their analytics and AI ambitions. However\, constructing a robust\, enterprise-grade data pipeline presents a myriad of challenges t hat organizations must learn to navigate to harness its full potential. It may seem daunting initially as it requires potentially high upfront inves tments\, often making it challenging to evaluate the necessary steps and R OI.\n\n### Intended Audience for the BoF\n- *Data Engineers\, Scientists\, and Architects:* Professionals responsible for building scalable\, compli ant data infrastructures. If you want to learn more about how to help your team build enterprise-grade data-driven products the talk is for you.\n- *IT Decision Makers (CTOs\, CIOs\, Business Leaders):* Leaders and executi ves seeking insights into strategic investments in data technology to maxi mize ROI leveraging data architecture.\n\n## Outline of the BoF\n![Enterpr ise-Ready Data at Scale](https://external-content.duckduckgo.com/iu/?u=htt p%3A%2F%2Fdrive.google.com/uc?id=1t9JcMieS6jE2TkqQmQJFPTXQsC_D3f0z)\n### C ore Components of an Enterprise Data Pipeline\n1. **Data Modeling:** Struc turing data in a way that it is usable and efficiently accessible. Creatin g and exposing internal vs customer-facing data models.\n\n2. **Overcoming Key Challenges:** \n - Methodologies to assess initial investments in technology and personnel against the expected returns\, emphasizing long-t erm value over short-term costs. \n - Ensuring data governance for comp liance such as DARE\, BYOK\, and working with UGC/PII data without stiflin g innovation.\n - Problems of Data Access for AI and Analytics:\n - *Data Silos*\n - *Data Quality Issues*\n - *Complexity in I ntegration*\n - *Access Permissions*\n\n4. **Rapid Innovation and Sc aling**\n - Powering customer data insights using Atlassian Analytics f or customers across different Atlassian products.\n - Data pipelines fo r model training required for AI/ML use cases.\n\n## Key Takeaway\n*Unlock Innovation Ensuring Compliance with High-Quality Advanced Data Management :* Real-world experience with tips and tricks from Atlassian in-house end- to-end data lifecycle management. High-quality\, accessible data is the c ornerstone of successful AI and analytics projects. The talk will help one understand various aspects to focus on while building data-products groun d up. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240705T110946Z LOCATION:Data engineering and infrastructure BOF area (Ground floor foyer) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/enterprise-ready-data- lifecycle-powering-ai-analytics-at-scale-SMcBc8YCPW7NFkB6fkSmaF BEGIN:VALARM ACTION:display DESCRIPTION:Enterprise-ready data life cycle: powering AI and analytics at scale in Data engineering and infrastructure BOF area (Ground floor foyer ) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T095000Z DTEND:20240713T100000Z DTSTAMP:20260419T104847Z UID:session/FqPKPjak3aGbHQRNb9kLMN@hasgeek.com SEQUENCE:3 CREATED:20240625T113809Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240703T055204Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in LLMs and GenAI track (Auditorium on the ground f loor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition DTSTART:20240713T095000Z DTEND:20240713T100000Z DTSTAMP:20260419T104847Z UID:session/TDZZFLEs4dmMjTMPnzyU9a@hasgeek.com SEQUENCE:3 CREATED:20240625T102023Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240704T102624Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition in Data engineering and infrastructure track (Semin ar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:From pair programming to piloting: building independent AI data an alysts DTSTART:20240713T100000Z DTEND:20240713T103500Z DTSTAMP:20260419T104847Z UID:session/XvrkcHNPykVgnJD791uQzM@hasgeek.com SEQUENCE:8 CATEGORIES:30 mins talk CREATED:20240625T113834Z DESCRIPTION:LLMs have transformed data analysis. With their ability to gen erate code to analyse data given appropriate prompts and instructions\, LL Ms are forming the bedrock of a new suite of data analysis tools.\n\nHowev er\, thanks to their ability to hallucinate\, and that they are occasional ly very poor logically\, the role of LLMs in data analysis so far has been restricted to being *copilots*\, who human data analysts can "pair progra m" with. \n\nThis talk is about how LLMs can go beyond being just copilots - how we can effectively pair LLMs with more conventional statistical met hods to produce true "AI Data analysts"\, that can generate insight and te ll stories from data without any human prompts. \n## Outline\n* Data analy sis capabilities of LLMs\n* Where LLMs fail at data analysis\n* How smart human analysts analyse data\n* Can LLMs plan? If they can't\, how can they be effective data analysts?\n* Putting together LLMs and statistical meth ods to generate data insight\n* Iterative reasoning and insight using a mu lti-agent framework\n* Using LLMs to tell effective data stories\n## Impac t\nAll of this is core to the product we are building at Babbage Insight\, with the objective of delivering Exceptional Insights to business decisio n-makers. \n\nThis approach will enable our product to anticipate question s that business decision-makers might have\, and present them with a perso nalised news feed of key stories about their business. GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T065402Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/building-an-ai-data-an alyst-XvrkcHNPykVgnJD791uQzM BEGIN:VALARM ACTION:display DESCRIPTION:From pair programming to piloting: building independent AI dat a analysts in LLMs and GenAI track (Auditorium on the ground floor) in 5 m inutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Apache XTables: translating things across tables DTSTART:20240713T100000Z DTEND:20240713T103500Z DTSTAMP:20260419T104847Z UID:session/MGssmSciNBmRyXcmodvn4E@hasgeek.com SEQUENCE:15 CATEGORIES:30 mins talk CREATED:20240628T050634Z DESCRIPTION:Apache Hudi\, Delta Lake\, and Iceberg are leading open-source projects that offer decoupled storage with transactional and metadata lay ers\, known as table formats in cloud storage. These formats store data in open columnar formats like Parquet and include metadata for schema\, comm it history\, partitions\, and column statistics. Selecting a table format can be challenging due to the unique features of each project. Enter XTabl e—an open-source project that ensures seamless interoperability between table formats. Instead of creating a new format\, XTable provides abstract ions for translating metadata\, enabling data to be written in any format and converted for use by various compute engines. This session will showca se XTable's solution to the challenges of format selection and interoperab ility in lakehouse workloads\, including a live demonstration of XTable in action. \nhttps://github.com/apache/incubator-xtable\n\nSpeaker: Vinish R eddy\nhttps://www.linkedin.com/in/vinish-reddy-pannala-868702108/ \n\nWho is the audience for this talk?\nData Engineers building data lake/lakehou se in their organisations. \n\nWhat is the problem you are trying to solve ?\nInter-operability across table formats\, catalogs and query engines. \ n\nWhat is the scope of this talk i.e.\, what content will you cover in th is talk?\nWe will start with intro about table formats and inter-operabili ty\, then deep dive into XTable. This will be followed by a live demo. In the end future road map for the OSS project will be discussed.\n\nHow will participants benefit from your talk?\nParticaptns can learn invaluable in sights in building data lakes and table formats\, hoping that it also exci tes them to contribute to Apache OSS projects.\n\n## Outline\n1. What are Table Formats ? What is Inter-Operability ? \n2. Intro to Apache XTable(In cubating) \n3. Deep dive on XTable and problems that can be solved using X Table. \n4. Demo and XTable in action.\n5. RoadMap and Future goals for th e OSS project. \n## Impact\nMore info can be found here. \nhttps://xtable. apache.org/ \nhttps://siliconangle.com/2023/11/15/onehouse-open-sources-on etable-data-tool-support-google-microsoft/ GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T082509Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/apache-xtable-incubati ng-interoperability-across-table-formats-MGssmSciNBmRyXcmodvn4E BEGIN:VALARM ACTION:display DESCRIPTION:Apache XTables: translating things across tables in Data engin eering and infrastructure track (Seminar halls on first floor) in 5 minute s TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Expert session - GenAI for Business DTSTART:20240713T100000Z DTEND:20240713T110000Z DTSTAMP:20260419T104847Z UID:session/NhUXod6Uh3cnLZaWRpLLMw@hasgeek.com SEQUENCE:2 CREATED:20240701T100709Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240712T065852Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Expert session - GenAI for Business in Board room (second floo r) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Evening break DTSTART:20240713T101500Z DTEND:20240713T104500Z DTSTAMP:20260419T104847Z UID:session/7n5uDhMZajxKpdQPijgDNS@hasgeek.com SEQUENCE:1 CREATED:20240701T095505Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T095512Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Evening break in Workshop\; BOFs for LLMs and GenAI track (Lib rary on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Evening break DTSTART:20240713T103500Z DTEND:20240713T105500Z DTSTAMP:20260419T104847Z UID:session/BtrMXqSzVvcCJdnnkYeFrb@hasgeek.com SEQUENCE:10 CREATED:20240625T115121Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240713T052722Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Evening break in LLMs and GenAI track (Auditorium on the groun d floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Evening break DTSTART:20240713T103500Z DTEND:20240713T110500Z DTSTAMP:20260419T104847Z UID:session/FowXrVExEPZyfNgH9MVTRU@hasgeek.com SEQUENCE:9 CREATED:20240625T103822Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240713T053201Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Evening break in Data engineering and infrastructure track (Se minar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Evening break DTSTART:20240713T104500Z DTEND:20240713T110000Z DTSTAMP:20260419T104847Z UID:session/PUuTyUmziMsfg2NNTpqXvc@hasgeek.com SEQUENCE:4 CREATED:20240705T110755Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240713T051054Z LOCATION:Data engineering and infrastructure BOF area (Ground floor foyer) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Evening break in Data engineering and infrastructure BOF area (Ground floor foyer) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Break DTSTART:20240713T110000Z DTEND:20240713T113000Z DTSTAMP:20260419T104847Z UID:session/299pz9AA9nhvTxLh8F9VC1@hasgeek.com SEQUENCE:2 CREATED:20240701T100726Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240701T100732Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Break in Board room (second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Wrap up session Track A: Data Engineering and Infrastructure DTSTART:20240713T110500Z DTEND:20240713T115000Z DTSTAMP:20260419T104847Z UID:session/74bRVGM5vFGnAbpRdomM2q@hasgeek.com SEQUENCE:16 CREATED:20240625T103855Z DESCRIPTION:In this wrap-up session\, editors and participants highlight s ome of the key takeaways from Track A: Data Engineering and Infrastructure at The Fifth Elephant 2024 Annual Conference GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071816Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Wrap up session Track A: Data Engineering and Infrastructure i n Data engineering and infrastructure track (Seminar halls on first floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:AMA on Building Apps with LLMs DTSTART:20240713T110500Z DTEND:20240713T122500Z DTSTAMP:20260419T104847Z UID:session/4e13hAstAYh3oVKC15dK4A@hasgeek.com SEQUENCE:6 CREATED:20240701T095838Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240713T051351Z LOCATION:Workshop\; BOFs for LLMs and GenAI track (Library on the second f loor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:AMA on Building Apps with LLMs in Workshop\; BOFs for LLMs and GenAI track (Library on the second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Sponsored talk: Accelerating data insights: building and enhancing relevance with GenAI by Amazon and DataStax DTSTART:20240713T110500Z DTEND:20240713T112500Z DTSTAMP:20260419T104847Z UID:session/FA5xiNsPhdJ1GBEDqViWyk@hasgeek.com SEQUENCE:5 CREATED:20240713T052549Z DESCRIPTION:In this session Amazon and DataStax partnered to explore the r apid development and enhancement of data relevance using Generative AI (Ge nAI). \n\nThe showcased how their collaboration has led to successful GenA I implementations\, delivering real-world results. During the session part icipants learnt how to leverage cutting-edge tools and techniques from AWS and DataStax to transform data-driven applications and achieve unparallel ed insights. \n\nThe session explored the future of intelligent data solut ions through practical examples and success stories from their joint effor ts.\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071910Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Sponsored talk: Accelerating data insights: building and enhan cing relevance with GenAI by Amazon and DataStax in LLMs and GenAI track ( Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Securing big data environments DTSTART:20240713T111500Z DTEND:20240713T121500Z DTSTAMP:20260419T104847Z UID:session/4EeQoR6Eox2FmQSn6z5ra3@hasgeek.com SEQUENCE:5 CATEGORIES:Birds of Feather (BOF) session CREATED:20240701T105327Z DESCRIPTION:\n\nThis BoF is about securing big data environments and learn ing about different controls from the security\, data privacy\, and compli ance side. How to balance security\, scale\, and user experiences while sc aling big data environment. There would be discussion around certain use c ases and edge cases that data platform team should be aware of while imple menting certain security controls \n\n\n\n## Who is the audience for your session?\n - Data Platform engineers/teams\n - Big data Infrastructure man agement engineers\n - Security engineers \n - GRC teams to help understand how to assess big data env.\n\n\n## What problem/pain are you trying to s olve (for the audience)?\n - The aim is to talk about critical security co ntrols on different big data env\n - To help data platform teams on how to go about implementing data security and data privacy controls on differen t big data env\n - Talk about certain open source tools that can be levera ged across different big data environments by data platform teams \n - How GRC and security teams can assess the data security posture for big data env.\n\n## What will be the scope of your session?\n - Real actionable ope n house discussion on how to secure big data env\n - challenges arising wh ile implementing certain security controls\n - How to implement data secur ity controls when you have already a lot of dashboards within your org bui lt on PII data. \n - Security with backward and forward compatibility in b ig data env\n - open source tools that can be leveraged in what kind of us e case\n\n## How will participants benefit from your session?\n - Real act ionable insights on how to secure big data env\n - understanding on how to manage different data privacy use cases\n - How to handle and manage comp liance (GDPR\, HIPAA etc)\n - Which controls can scale and how and what ar e different gotchas that one has to be aware of GEO:12.9666826;77.6352903 LAST-MODIFIED:20240713T043840Z LOCATION:Data engineering and infrastructure BOF area (Ground floor foyer) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/securing-big-data-envi ronments-4EeQoR6Eox2FmQSn6z5ra3 BEGIN:VALARM ACTION:display DESCRIPTION:Securing big data environments in Data engineering and infrast ructure BOF area (Ground floor foyer) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Transition session DTSTART:20240713T112500Z DTEND:20240713T113500Z DTSTAMP:20260419T104847Z UID:session/GZxofjCEENxCaGMT6DkV6f@hasgeek.com SEQUENCE:2 CREATED:20240713T052641Z GEO:12.9666826;77.6352903 LAST-MODIFIED:20240713T052722Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Transition session in LLMs and GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:GenAI demo showcases DTSTART:20240713T113000Z DTEND:20240713T123000Z DTSTAMP:20260419T104847Z UID:session/Lua6GVNgHJnx4T2vdZM2XQ@hasgeek.com SEQUENCE:2 CREATED:20240701T100746Z DESCRIPTION:**The GenAI Demo showcase is presented by The Fifth Elephant c ommunity in partnership with [E2E Networks Limited](https://www.e2enetwork s.com/).**\n\nFor more information about participants in the showcase sess ion check https://hasgeek.com/fifthelephant/genai-demo-showcase/ GEO:12.9666826;77.6352903 LAST-MODIFIED:20240712T080121Z LOCATION:Board room (second floor) - Bangalore International Centre\nBanga lore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:GenAI demo showcases in Board room (second floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Wrap up session Track B - GenAI and LLMs in Business DTSTART:20240713T113500Z DTEND:20240713T115500Z DTSTAMP:20260419T104847Z UID:session/RWh53VyybWfsPQE8d4Vob4@hasgeek.com SEQUENCE:12 CREATED:20240625T115142Z DESCRIPTION:In this wrap-up session\, editors and participants highlight s ome of the key takeaways from Track B - GenAI and LLMs in Business at The Fifth Elephant 2024 Annual Conference GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071808Z LOCATION:LLMs and GenAI track (Auditorium on the ground floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Wrap up session Track B - GenAI and LLMs in Business in LLMs a nd GenAI track (Auditorium on the ground floor) in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Book Discussion on Dream Machine: A Graphic Novel about AI DTSTART:20240713T115000Z DTEND:20240713T123500Z DTSTAMP:20260419T104847Z UID:session/HiEYddS3NgGkScc9tJERft@hasgeek.com SEQUENCE:11 CATEGORIES:Birds of Feather (BOF) session CREATED:20240701T105412Z DESCRIPTION:Join us for a candid chat with the artist Appupen\, whose rece ntly released graphic novel Dream Machine explores the implications of unl eashing AI to the real world. Through the narrative centered around Hugo — an entrepreneur who dreams of being a superhero — the novel uncovers some crucial concerns around implementing AI at scale such as Bias\, Surv eillance\, Ethics\, Trust & Creativity.\n\nThis discussion will be facilit ated by Rasagy Sharma.\n\n## Outline\n* Introduction to Appupen: Past work & key themes\n* Introduction to Dream Machine: Context & Plot\n* Open dis cussion on some of the themes around AI’s impact\n* Concluding remarks o n the future of AI & creativity\n* Open Q&A with the artist\n\n## Tips for attending\n* We recommend you to read the novel if you wish to engage in the conversations around the narrative\n* You’re welcome to join the ses sion even if you haven’t read the novel\, and contribute on the broader AI themes.\n* Limited copies of Dream Machine will be on sale at the event .\n\n## References\n* [Excerpt from Dream Machine](https://scroll.in/artic le/1063410/springtime-in-ai-land-appupens-new-graphic-novel-explores-the-a dvent-of-artificial-intelligence)\n* [In-depth review of Dream Machine](ht tps://scroll.in/article/1068209/appupens-new-graphic-novel-dream-machine-a sks-difficult-questions-of-artificial-intelligence)\n* [More work from App upen](https://appupen.wordpress.com/)\n* [Appupen on Instagram](https://ww w.instagram.com/appupen/)\n GEO:12.9666826;77.6352903 LAST-MODIFIED:20240816T071329Z LOCATION:Data engineering and infrastructure track (Seminar halls on first floor) - Bangalore International Centre\nBangalore\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/2024/schedule/book-discussion-on-dre am-machine-a-graphic-novel-about-ai-HiEYddS3NgGkScc9tJERft BEGIN:VALARM ACTION:display DESCRIPTION:Book Discussion on Dream Machine: A Graphic Novel about AI in Data engineering and infrastructure track (Seminar halls on first floor) i n 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT END:VCALENDAR