Efficient Agentic Systems: Fireside Chat + Happy Hour

​Your agents burn tokens. Let’s talk about making them as cheap as possible.

​Join us for a conversation with Ryan Wang, SVP of Applied AI at Nebius, and Kyle Kranen, Engineering Leader for NVIDIA Dynamo, on how choices across hardware, inference, and models affect the speed, cost, quality, and efficiency of agentic systems—and what builders need to get right when agents move from prototype to production.

​No decks. No pitch. Just a candid technical conversation, followed by food, drinks, and networking with people building the stack agents run on.

​In San Jose for PyTorch? Join us at The Pressroom, less than a mile from the conference center, for a conversation that goes beyond the conference agenda.

​What to expect

• ​An honest conversation with leaders working across AI infrastructure, inference, and models

• ​Practical lessons and tradeoffs from moving agentic systems into production

• ​What’s working, what isn’t, and what builders are learning along the way

• ​A room of engineers, founders, and technical leaders putting AI into production

• ​Food, drinks, and plenty of time to meet other builders

​Who should come

• ​AI and ML engineers, developers, technical founders, researchers, and infrastructure teams shipping agents beyond the demo stage

• ​Builders working anywhere in the stack—from hardware, models, and inference to orchestration, evaluation, and applications

• ​Anyone who wants to learn from people solving real production challenges

• ​People who believe “AI in production” should mean more than calling an LLM once

​Agenda

• ​6 - 6:30PM — Welcome, Drinks & Networking

• ​6:30 - 7PM — Fireside chat + Q&A

• ​7 - 8PM — Food, drinks, and networking

​Speakers

​Nebius – Ryan Wang, SVP Applied AI

​NVIDIA – Kyle Kranen, Engineering Leader for NVIDIA Dynamo

​Small room. Registration approval is required.

​About Nebius 
Nebius, the AI cloud company, is building the full-stack platform for developers and companies to take charge of their AI future, from data and model training to production deployment. Founded on deep in-house technological expertise and operating at scale with a rapidly expanding global footprint, Nebius serves startups and enterprises building AI products, agents and services worldwide.

​Nebius is listed on Nasdaq (Nasdaq: NBIS) and headquartered in Amsterdam.

​For more information, visit www.nebius.com.