BEGIN:VCALENDAR VERSION:2.0 PRODID:-//HasGeek//NONSGML Funnel//EN DESCRIPTION:A workshop on developing skills to read ML-Sys papers X-WR-CALDESC:A workshop on developing skills to read ML-Sys papers NAME:Beyond the jargon: making research papers work in the real world X-WR-CALNAME:Beyond the jargon: making research papers work in the real wo rld REFRESH-INTERVAL;VALUE=DURATION:PT12H SUMMARY:Beyond the jargon: making research papers work in the real world TIMEZONE-ID:Asia/Kolkata X-PUBLISHED-TTL:PT12H X-WR-TIMEZONE:Asia/Kolkata BEGIN:VEVENT SUMMARY:Beyond the jargon: making research papers work in the real world DTSTART:20250718T121500Z DTEND:20250718T140000Z DTSTAMP:20260421T100334Z UID:session/7DQPRgmRY5f92yTKu5F3ZF@hasgeek.com SEQUENCE:21 CREATED:20250708T034724Z DESCRIPTION:## About the workshop\nThis workshop is designed to bring toge ther **engineers\, researchers\, and technologists** at the intersection o f **machine learning and systems** to learn how to effectively read\, unde rstand\, and apply research papers in real-world scenarios.\n\nThe goal is not to passively read or summarize academic content\, but to **actively d issect the paper\, understand its motivation\, techniques\, and trade-offs ** - just like we do when building or adapting a system ourselves.\n\n---\ n\n## Takeaways for participants\n- Equip participants with a **framework to approach applied research papers**.\n- Demonstrate how to **extract reu sable techniques and mental models** from academic work.\n- Encourage **cr itical thinking around performance bottlenecks\, architectural trade-offs\ , and design decisions** as outlined in state-of-the-art research.\n- Fost er a **community of engineers and researchers passionate about production- grade ML systems**.\n\n---\n\n## About the paper + session format\nWe'll b e discussing ["Bullion: A Column Store for Machine Learning"](https://arxi v.org/pdf/2404.08901v3). \n\nBullion is a next-gen columnar storage system tailored specifically for machine learning workloads\, offering specializ ed support for deletion-compliance\, long-sequence sparse features\, and f eature quantization. It optimizes wide-table metadata parsing and enables efficient multimodal and vector-based ML data handling\, significantly red ucing I/O costs and storage overhead compared to traditional column stores . \n\nThis is a **deeply systems-oriented paper with real-world implicatio ns**\, a perfect start to demonstrate:\n- How to **extract implementation ideas** from a paper \n- How to **reason about tradeoffs** \n- How to ** identify assumptions and possible extensions**\n\nWe won’t just _“go t hrough”_ the paper. Instead\, we’ll critically explore:\n- What proble m is being solved? \n- What are the design decisions? \n- Could this app ly in our context and/or day jobs?\n\n---\n\n## What participants need to do or know\nThis is a **hands-on reading workshop\, not a lecture**.\n\n** Before the session:**\n\n- Read the **abstract\, introduction\, and conclu sion** of the paper\n- Think about: _What do you understand about the prob lem space? What would you do differently?_\n- Bring a **notepad or laptop* * to jot down observations\n- *Optional but encouraged:* skim through how **Parquet file formats** work and typical **ML data loading pipelines**\n\ n**Ideal for folks who:**\n- Work at the **systems/ML boundary**\n- Are in terested in **research-to-production translation**\n- Want to **sharpen th eir research reading and technical reasoning skills**\n\n---\n\n## About t he facilitators\n**Aditi Ahuja** \nCurrently working in the **Search team at Couchbase**\, focused on full text and vector search. Has been a speak er at **Fifth Elephant\, Gophercon India\, and PromDay (a Kubecon co-locat ed event)**. Past internships include an **LFX mentorship at the Thanos pr oject (a CNCF project).**\n\n**Harini Anand** \n**CSE undergrad** passion ate about **Computational Cognition\, ML\, and AI in Healthcare**. **SDE I ntern at IBM Data & AI**\, working on watsonx™. Formerly at **Niramai & IIT Hyderabad**\, researching ML for breast cancer and gene regulatory net works. Built cognitive tools for dementia prevention as a student entrepre neur. **Google KaggleX Mentee\, AWS Scholar\, Harvard WE Tech Fellow\, and Oxford & MIT Summer School alumna**. Advocate for **STEM representation** \, speaker\, and published AI researcher.\n\n**Abhinav Upadhyay**\nIndepen dent systems engineer who explores the internals of software and hardware through his writing. He publishes [Confessions of a Code Addict](https://b log.codingconfessions.com/)\, a newsletter focused on compilers\, interpre ters\, operating systems\, and performance engineering. With over a decade of experience in backend systems and machine learning\, he enjoys diving deep into how things work and sharing those insights with fellow engineers .\n\n## About the organizers\n[**Bengaluru Systems Meetup**](https://hasge ek.com/bengalurusystemsmeetup) brings together Bengaluru’s systems enthu siasts in meetups that cover Databases\, Distributed Systems\, Compilers\, Orchestration Systems\, and Dataflow Systems.\n\n[**The Fifth Elephant**] (https://hasgeek.com/fifthelephant) is a community of practitioners\, who share feedback on data\, AI and ML practices in the industry. \n\n💬 **P ost a comment** with your questions here.\n\n## Contact information\n📞 Call The Fifth Elephant at (91) 7676332020\n📧 For inquiries about regis tration\, drop an email to info@hasgeek.com LAST-MODIFIED:20250713T143037Z LOCATION:To be announced - https://hasgeek.com/fifthelephant/making-resear ch-papers-work-in-real-world-workshop/ ORGANIZER;CN="The Fifth Elephant":MAILTO:no-reply@hasgeek.com URL:https://hasgeek.com/fifthelephant/making-research-papers-work-in-real- world-workshop/ BEGIN:VALARM ACTION:display DESCRIPTION:Beyond the jargon: making research papers work in the real wor ld in 5 minutes TRIGGER:-PT5M END:VALARM END:VEVENT END:VCALENDAR