SF #TechWeek: Data & AI Edition

​Join us for SF Tech Week: Data & AI Edition, hosted by BentoML, Datastrato, and Zilliz. This event will focus on the latest developments in AI and data technologies, featuring talks from representatives of each hosting company. It’s a great opportunity to learn, share ideas, and connect with professionals in the field. Whether you're an expert or just curious about AI, this meetup offers valuable insights and networking opportunities.

​Agenda:
5:30 - 6:00 - Welcome/Networking/Registration
6:00 - 6:25 - Multimodal Search with Open-Source Tools, Stefan Webb, DevRel, Zilliz
6:25 - 6:50 - Practical Guide to Deploying LLMs, Chaoyu Yang, Founder & CEO, BentoML
7:00 - 7:30 - One single binary to tackle streaming and historical analytics, Ken Chen - Co-Founder & Chief Architect, Timeplus
7:30 - 7:40 - Enhancing Context Provision for LLMs: Beyond RAG and Prompts, Brian Hart, Engineer, Tecton
7:40 - 8:30 - Networking

​Tech talk 1: Multimodal Search with Open-Source Tools
Speaker: Stefan Webb, DevRel, Zilliz
Abstract: A recent and exciting development in the world of Generative AI has been the use of language to understand images, video, and sound. One example is multi-modal retrieval, which is the process of using one modality, like text, to search another modality, like images. It is not only useful for search engines across media types, but also for grounding LLMs in factual data and reducing hallucinations. In this talk, I explain how to build a simple but performant multi-modal retrieval pipeline using completely open-source tools and models: the vector database Milvus and HuggingFace libraries for modeling and data. I discuss techniques to use multimodal retrieval most effectively and increase recall, as well as some interesting and diverse industry applications.

​Tech Talk 2: Practical Guide to Deploying LLMs
Speaker: Chaoyu Yang, Founder & CEO, BentoML

​Tech Talk 3: Timeplus: One single binary to tackle streaming and historical analytics
Speaker: Ken Chen, Co-Founder/Chief Architect, Timeplus

Lightning talk: Enhancing Context Provision for LLMs: Beyond RAG and Prompts
Speaker: Brian Hart, Engineer, Tecton
Abstract: What is the most effective way to provide context to LLMs? Practitioners often emphasize two main strategies: building vector databases and crafting "optimal" prompts. While both are crucial for the success of Generative AI applications, they have inherent limitations. Vector search relies on similarity, which isn't always the best method for information retrieval. Conversely, incorporating all available context into a single prompt is neither effective nor efficient.
In this talk, we will examine various scenarios to illustrate why traditional prompt engineering and Retrieval-Augmented Generation (RAG) approaches fall short. We will explore additional methods for providing context in GenAI applications. Furthermore, we will demonstrate how to leverage Tecton's new GenAI features to rapidly build and test an agentic workflow.

​This event is a part of #TechWeek - a week of events hosted by VCs and startups to bring together the tech ecosystem