AI/ML Infra Meetup with OpenAI, Poshmark & Alluxio
Important update (August 28th): Our in-person event has reached capacity. If you are currently on the waitlist or your registration is pending approval, we regrettably cannot guarantee you a spot at the venue. However, we encourage you to participate virtually using the following Zoom link: https://us06web.zoom.us/webinar/register/WN_ZfoTN38MSKOsOfoGrKZHvg
Join the AI/ML Infra Meetup at Poshmark's headquarters in Redwood City, CA, co-hosted by Alluxio and Poshmark. Hear from a great lineup of speakers, including OpenAI, Poshmark, Alluxio, and more, who will share invaluable insights on building AI infrastructure at scale.
If you're passionate about the latest advancements in AI/ML infrastructure, including LLMs, vector databases, training & fine-tuning techniques, RAG, GPU optimization, PyTorch, and more, then this is the event for you!
Immerse yourself in learning, networking, and conversations. In the end, enjoy the mix-and-mingle happy hour. Food and drinks are on us!
*Note: This registration is for in-person attendance only. For virtual participation, please use this Zoom link: https://us06web.zoom.us/webinar/register/WN_ZfoTN38MSKOsOfoGrKZHvg
Agenda (in PDT)
Door opens at 4:00 pm
4:00 pm: Check-in and Happy Hour
5:00 pm: Welcome Remarks
5:05 pm: Preference Tuning and Fine Tuning LLMs (Ankit Khare @ OpenAI)
5:30 pm: Maximizing GPU Efficiency: Optimizing Model Training with GPU Anywhere (Bin Fan @ Alluxio)
5:55 pm: Scaling Vector Databases for E-Commerce Visual Search: Architectural Strategies for 150M+ Product Search & Recommendations (Mahesh Pasupulet @ Poshmark)
6:10 pm: Scaling Experimentation Platform in Digital Marketplaces: Architecture, Implementation & Lessons Learned (Koundinya Pidaparthi @ Poshmark)
6:30-7:30 pm: Networking Happy Hour continued
Join the Alluxio Community
We are hiring! Check out the job openings
🧑💻 Join Alluxio Community Slack
🐦 Follow Alluxio on Linkedin & Twitter / X
📺 Subscribe to the Alluxio YouTube Channel
📖 Download the PyTorch Tuning Guide
Speakers
Ankit Khare, Developer Experience Engineer @ OpenAI
Bin Fan, VP of Technology @ Alluxio
Mahesh Pasupuleti, VP of DS, ML & Data Infra @ Poshmark
Koundinya Pidaparthi, VP of Analytics @ Poshmark
Presentation Details
Talk 1: Preference Tuning And Fine Tuning LLMs
This session aims to provide practical insights for AI enthusiasts on effectively customizing and leveraging LLMs in various applications through preference tuning and fine-tuning.
About the speaker: Ankit Khare is currently a Developer Experience Engineer at OpenAI. He is passionate about AI and building developer communities. He has helped build developer relations functions at several startups, including Rockset, Twelve Labs, and Abacus.AI.
Talk 2: Maximizing GPU Efficiency: Optimizing Model Training with GPU Anywhere
In the rapidly evolving landscape of AI and machine learning, infra teams face critical challenges in managing large-scale data for AI. Performance bottlenecks, cost inefficiencies, and management complexities pose significant challenges for AI platform teams supporting large-scale model training and serving.
In this talk, Bin Fan, VP of Technology @ Alluxio, will discuss the challenges of I/O stalls that lead to suboptimal GPU utilization during model training. He will present a reference architecture for running PyTorch jobs with Alluxio in cloud environments, demonstrating how this approach can significantly enhance GPU efficiency.
What you will learn:
How to identify GPU utilization and I/O-related performance bottlenecks in model training
Leverage GPU anywhere to maximize resource utilization
Best practices for monitoring and optimizing GPU usage across training and serving pipelines
Strategies for reducing cloud costs and simplifying management of AI infrastructure at scale
Talk 3: Scaling Vector Databases for E-Commerce Visual Search: Architectural Strategies for 150M+ Product Search & Recommendations
In the rapidly evolving world of e-commerce, visual search has become a game-changing technology. Poshmark, a leading fashion resale marketplace, has developed Posh Lens – an advanced visual search engine that revolutionizes how shoppers discover and purchase items.
Under the hood of Posh Lens lies Milvus, a vector database enabling efficient product search and recommendation across our vast catalog of over 150 million items. However, with such an extensive and growing dataset, maintaining high-performance search capabilities while scaling AI infrastructure presents significant challenges.
In this talk, Mahesh Pasupuleti, VP of DS, ML & Data Infra @ Poshmark, will share:
The architecture and strategies to scale Milvus effectively within the Posh Lens infrastructure
Key considerations include optimizing vector indexing, managing data partitioning, and ensuring query efficiency amidst large-scale data growth
Distributed computing principles and advanced indexing techniques to handle the complexity of Poshmark's diverse product catalog
Talk 4: Scaling Experimentation Platform in Digital Marketplaces: Architecture, Implementation & Lessons Learned
Scaling experimentation in digital marketplaces is crucial for driving growth and enhancing user experiences. However, varied methodologies and a lack of experiment governance can hinder the impact of experimentation, leading to inconsistent decision-making, inefficiencies, and missed opportunities for innovation. At Poshmark, we developed a homegrown experimentation platform, Lightspeed, that allowed us to make reliable and confident reads on product changes, which led to a 10x growth in experiment velocity and positive business outcomes along the way.
In this session, Koundinya Pidaparthi, VP of Analytics @ Poshmark, will discuss the best practices and lessons learned from successful large-scale experiment implementations. He will explore the importance of experimentation, overcome scalability challenges, and gain insights into the frameworks and technologies that enable effective testing.
You can check what the past events were like through:
If you have any questions, please contact Hope Wang at hope.wang@alluxio.com.
Call for speakers! We are seeking passionate individuals as speakers. If you have valuable insights to share with the AI/ML Infra community, please fill out this form: https://forms.gle/iJX9GTMaAVQdzKc28.