Personalizing Long-running Agents

Friday, Oct 9

4:00pm – 7:00pm

PTET

Hosted by

-‿-

⚆◟⚆

302/350 spots left

This is an in-person event! Registration is required in order to get in. ​Topic: Sonoma AI Meetups for remote tech workers, to build a North Bay AI community, collaboratively learn, share ideas, and network with others who are automating AI Agents. ​4:00 - Doors open/Checkin 4:25 - Welcome 4:30–4:55 PM — Teaching Claude and Codex to Think in GPU: Agentic Skills for CUDA-X Data Science 5:00–5:25 PM — Personalizing Long-running AI Agents ​5:30–6:30 PM — Breakout Sessions 6:30-7:00 PM – Networking BREAKOUT ROOM 1 — Intro to LLMs & Agents + Open Q&A · Learn the basics and bring your questions · No signup required. BREAKOUT ROOM 2 — Lightning Pitches Present your AI project, idea, or demo in 9-minutes. 6 presenter spots · Signup required https://bit.ly/4zK0m30 BREAKOUT ROOM 3 — AI Expert Office Hours Reserve 15 minutes of 1:1 help with an AI expert to get unstuck · Signup required https://bit.ly/3U9Kruv · Office Hour Topics include: • AI agents & applications • Full-stack data science • Fine-tuning LLMs Office Hour Experts include Allison Ding (NVIDIA), Russell Horton (GitHub Next), and others. Tech Talk with Demo #1: Teaching Claude and Codex to Think in GPU: Agentic Skills for CUDA-X Data Science Speaker: Allison Ding https://www.linkedin.com/in/allison-ding/ Description: Agentic skills turn a generic AI assistant into a domain-aware agent that selects the right tool, generates validated code, and catches its own failure modes. This talk shows how to use NVIDIA's CUDA-X data science skills with Claude, Codex, and OpenClaw to generate correct, GPU-accelerated code without deep CUDA-X expertise — and pulls back the curtain on how they work: trigger conditions, library selection rules, documentation grounding, and runtime validation that prevents silent CPU fallback. The second half is a live demo: install the skills from NVIDIA's GitHub catalog with a single CLI command, then watch agents accelerate a pandas pipeline with cuDF, train a classifier with cuML, and build a nearest-neighbor index with cuVS — each prompt routed to the right GPU path automatically. Tech Talk with Demo #2: Personalizing Long-running AI Agents Speaker: Russell Horton https://www.linkedin.com/in/linguistics/ ​Where: This is an in-person event. Registration using this form is required to get into the event. Thank you to our venue sponsor SOMO Cowork https://www.somocowork.com Are you interested in speaking or sponsoring a Sonoma AI event? 🍷 Let us know! https://forms.gle/CEFNf844CbUiVPVP7 Watch talks from our last event: https://www.youtube.com/@sonoma-ai Hotels: Staying overnight instead of a long round trip drive? All of these ~15min drive, except Bodega Bay is a little longer. ​• DoubleTree by Hilton Hotel Sonoma Wine Country a hub for the Sonoma airporter bus; quiet; lots of golfing! ​• Hotel Centro Sonoma Wine Country, Tapestry by Hilton pricing is typically quite good; newly constructed ​• Hotel La Rose, Santa Rosa funky little historical place; great deli downstairs; in the hip-ish 4th Street district ​• Courtyard Santa Rosa reliably good; friends from SF tend to stay here ​• Hyatt Regency Sonoma Wine Country similar to Courtyard above ​• Lodge at Bodega Bay, beautiful views; 25 minute drive from Sebastopol This event is a part of #SFTechWeek—a week of events hosted by VCs and startups to bring together the tech ecosystem. Learn more at www.tech-week.com.

Guest List

48 Going · 1 Interested · 2 Maybe

View all

˜▾˜

*⏖*

❛ᴗ❛