BEGIN:VCALENDAR VERSION:2.0 PRODID:-//HasGeek//NONSGML Funnel//EN DESCRIPTION:Conference X-WR-CALDESC:Conference NAME:Practicing MLOps in your organization: tools\, frameworks and governa nce X-WR-CALNAME:Practicing MLOps in your organization: tools\, frameworks and governance REFRESH-INTERVAL;VALUE=DURATION:PT12H SUMMARY:Practicing MLOps in your organization: tools\, frameworks and gove rnance TIMEZONE-ID:Asia/Kolkata X-PUBLISHED-TTL:PT12H X-WR-TIMEZONE:Asia/Kolkata BEGIN:VEVENT SUMMARY:Registration and check-in DTSTART:20221111T040000Z DTEND:20221111T043000Z DTSTAMP:20260311T230035Z UID:session/6KRB6dFp1vdDFhJ1g7Dbn1@hasgeek.com SEQUENCE:9 CREATED:20221020T054220Z GEO:12.9661806;77.5963804 LAST-MODIFIED:20221102T092122Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Registration and check-in in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Introduction to MLOps Conference\; code of conduct and schedule DTSTART:20221111T043000Z DTEND:20221111T044500Z DTSTAMP:20260311T230035Z UID:session/NNz3hFLRWVQGp65WiXkVUz@hasgeek.com SEQUENCE:7 CREATED:20221020T054239Z GEO:12.9661806;77.5963804 LAST-MODIFIED:20221109T083457Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Introduction to MLOps Conference\; code of conduct and schedul e in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Panel: Developing ML applications DTSTART:20221111T044500Z DTEND:20221111T053000Z DTSTAMP:20260311T230035Z UID:session/RWVo6jCacdh8XES25HNVTU@hasgeek.com SEQUENCE:36 CREATED:20221020T054642Z DESCRIPTION:There has been significant investment in data\, teams\, and in frastructure\, but the impact and adoption has been incremental. The purpo se of this chat is to discuss ways to improve adoption of ML - systematic identification and scoping of use cases\, organizational alignment and coo peration\, and effective management of the delivery process.\n\nThis panel is sponsored by Scribble Data. GEO:12.9661806;77.5963804 LAST-MODIFIED:20221219T081112Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Panel: Developing ML applications in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Fast development of ML applications - do the right thing. DTSTART:20221111T053000Z DTEND:20221111T060500Z DTSTAMP:20260311T230035Z UID:session/JrTK3AsCB2dA2arx2TWVUU@hasgeek.com SEQUENCE:41 CREATED:20221028T044649Z DESCRIPTION:# Speaker\n\nVenkata Pingali\, Co-Founder & CEO\, Scribble Dat a\n\n# Abstract\n\nWidespread adoption of machine learning (ML) in industr y is still a challenge today due to resource constraints and RoI questions . Production ML approaches today require high skill\, rely on large volume s of data\, and have long delivery timelines. In this talk\, we argue for Sub-ML - a class of ML simpler than traditional ML approaches\, often desi gned to be used in decision support systems\, and delivered under tight co nstraints. Sub-ML\, also called as ML-at-reasonable-scale (MLRS) and Analy tical ML\, covers upto 80% of the ML usecases in an enterprise. Characteri zed by their speed in realizing business value and support for diverse use cases\, Sub-ML applications still require guarantees of correctness\, tra nsparency\, and auditability in the data transformation process. We draw o n our experience in the fin-tech\, ed-tech and e-commerce domains to lay o ut design choices for feature stores to enable Sub-ML\, tradeoffs we made including constraining the problem space\, bundling capabilities for fast development\, and incorporating a data consumption layer.\n\nThis talk is sponsored by Scribble Data. GEO:12.9661806;77.5963804 LAST-MODIFIED:20230810T072606Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/mlops-2022/schedule/do-the-right-thi ng-fast-development-of-ml-applications-using-sub-ml-feature-stores-JrTK3As CB2dA2arx2TWVUU BEGIN:VALARM ACTION:display DESCRIPTION:Fast development of ML applications - do the right thing. in R eactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Break DTSTART:20221111T060500Z DTEND:20221111T062000Z DTSTAMP:20260311T230035Z UID:session/F7be51HJSxEyQFcNpWapoK@hasgeek.com SEQUENCE:22 CREATED:20221028T044729Z LAST-MODIFIED:20221109T083414Z LOCATION:Bengaluru\, India ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Break in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:How PayPal uses Inference Graphs to design ML systems DTSTART:20221111T062000Z DTEND:20221111T065500Z DTSTAMP:20260311T230035Z UID:session/QhiWqPTKFg49K3p24YWKm4@hasgeek.com SEQUENCE:35 CREATED:20221028T044750Z DESCRIPTION:# Speakers\n\nSaurav Raj (ML Engineer\, PayPal)\nSharmili Srin ivasan (ML Engineer\, PayPal)\n\n# Abstract\n\nMachine learning inference pipelines are getting complex as they often comprise multiple models to ma ke a prediction. These models typically follow patterns such as chaining\, fanout\, or ensemble in production. A streamlined model deployment mechan ism that supports these patterns would reduce the effort of ML Engineers a nd make the ML systems more agile. In this talk\, we will learn about infe rence graphs in ML systems and use them to design and deploy ML pipelines that require composing multiple ML models. GEO:12.9661806;77.5963804 LAST-MODIFIED:20230810T072606Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/mlops-2022/schedule/back-to-the-futu re-designing-multi-model-machine-learning-systems-using-inference-graphs-Q hiWqPTKFg49K3p24YWKm4 BEGIN:VALARM ACTION:display DESCRIPTION:How PayPal uses Inference Graphs to design ML systems in React or room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:War stories on MLOps and data governance - learnings from Data Mes h DTSTART:20221111T065500Z DTEND:20221111T073000Z DTSTAMP:20260311T230035Z UID:session/J5sB936JcNh9qMWQo5g3GW@hasgeek.com SEQUENCE:28 CREATED:20221028T050318Z DESCRIPTION:The operation\, process workflow and automation for ML models are challenges for any organization. The transparency and visibility of th ese models across ecosystems are important to provide explanations on info rmed decisions. In order to conform to regulations and meet stakeholder ex pectations\, it is also important to use the right operating models and pr ocesses. In this talk\, we will learn about our journey in building ML use cases and the typical challenges faced in ML productionisation. The diffe rent processes that can be automated for consistency will also be explored .\n\nThis talk is sponsored by Data Mesh. GEO:12.9661806;77.5963804 LAST-MODIFIED:20221219T075834Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:War stories on MLOps and data governance - learnings from Data Mesh in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Lunch DTSTART:20221111T073000Z DTEND:20221111T083000Z DTSTAMP:20260311T230035Z UID:session/FYuT78i7NnReVjJdCeDBiq@hasgeek.com SEQUENCE:16 CREATED:20221020T054542Z GEO:12.9661806;77.5963804 LAST-MODIFIED:20221109T081706Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Lunch in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Panel: Building value to MLOps with data governance DTSTART:20221111T083000Z DTEND:20221111T091500Z DTSTAMP:20260311T230035Z UID:session/RPWHJpTxPF4orduQ7cRi1W@hasgeek.com SEQUENCE:26 CREATED:20221020T054620Z DESCRIPTION:This Panel Discussion is sponsored by ThoughtWorks. GEO:12.9661806;77.5963804 LAST-MODIFIED:20230102T045308Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Panel: Building value to MLOps with data governance in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Make AI real - from labs to production DTSTART:20221111T091500Z DTEND:20221111T095000Z DTSTAMP:20260311T230035Z UID:session/3cfv57tPfQxZhTWQwfe7G8@hasgeek.com SEQUENCE:23 CREATED:20221031T054741Z DESCRIPTION:# Speaker\n\nRavi Kumar Meduri (Executive Vice President - IP and Digital Solutions)\n\n# Abstract\n\nThe organization and management of the lifecycle of data is foundational for AI. It is imperative that the v alue is in democratizing data and letting the business analysts\, research ers and innovators in the enterprise use it for more effective decision ma king. This unification of data and scaling of AI/ML practices requires gov ernance and an operationalization framework broadly named as MLOps. In thi s talk\, we will take a fresh perspective on MLOps\, its process\, tooling \, governance\, and also deep dive into Explainable AI\, Responsible AI an d incorporate them into MLOps.\n GEO:12.9661806;77.5963804 LAST-MODIFIED:20230810T072606Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/mlops-2022/schedule/2022-an-operatio nalisation-odyssey-of-ai-from-labs-to-production-3cfv57tPfQxZhTWQwfe7G8 BEGIN:VALARM ACTION:display DESCRIPTION:Make AI real - from labs to production in Reactor room in 5 mi nutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:ML at scale: how Udaan built its ML platform DTSTART:20221111T095000Z DTEND:20221111T102500Z DTSTAMP:20260311T230035Z UID:session/93dWF65sSZYacqbtRvFdar@hasgeek.com SEQUENCE:26 CREATED:20221028T044710Z DESCRIPTION:# Speakers\n\nDr. Mohit Kumar (VP & Head of Data Science\, Dat a Platform and Product Analytics\, Udaan)\n\nSai Sharan Tangeda (Product E ngineer of Data\, ML & Infrastructure\, Udaan)\n\nSampan Nayak (Engineer D ata & ML Platform\, Udaan)\n\n# Abstract\n\nThe MLOps initiatives at Udaan will be introduced along with the different components that power its ML Platform. We will focus on the approaches that we have taken and the tooli ng choices that were made to systemize MLOps to operate at scale. The diff erent tools that were evaluated\, and the challenges faced over time will be shared in the talk. All the learning experience eventually led us in cr eating a refined ML platform which includes Feature Store\, Pipelines and Deploy. GEO:12.9661806;77.5963804 LAST-MODIFIED:20230810T072606Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/mlops-2022/schedule/come-and-see-ml- at-scale-unified-ml-platform-at-udaan-93dWF65sSZYacqbtRvFdar BEGIN:VALARM ACTION:display DESCRIPTION:ML at scale: how Udaan built its ML platform in Reactor room i n 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT BEGIN:VEVENT SUMMARY:Closing session - takeaways DTSTART:20221111T102500Z DTEND:20221111T103000Z DTSTAMP:20260311T230035Z UID:session/QLCCbu7jJZeoJi7PnYVLwR@hasgeek.com SEQUENCE:11 CREATED:20221020T054658Z GEO:12.9661806;77.5963804 LAST-MODIFIED:20221109T082007Z LOCATION:Reactor room - Microsoft Reactor\nBengaluru\nIN ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com BEGIN:VALARM ACTION:display DESCRIPTION:Closing session - takeaways in Reactor room in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT END:VCALENDAR