vLLM x Crusoe Meetup: Production Open Source Inference

​Calling all AI engineers, researchers, and infrastructure builders: Join us for the vLLM x Crusoe meetup in San Francisco!

​This is a unique opportunity to connect with vLLM maintainers and the Crusoe engineers architecting high-performance GPU clusters for production-scale open-weights workloads. See the powerful synergy between the vLLM ecosystem with Crusoe’s upstream contribution and application throughout its cloud Infrastructure.

​The session will dive deep into the frontier of inference. The vLLM team's Zijing Liu and Zach Xi will share technical updates on the vLLM roadmap, serving for agentic workloads, and advanced KV cache offloading techniques. Crusoe's Omer Landau and Eesha Pathak will detail the infrastructure powering these workloads, from NVIDIA and AMD GPU orchestration to MemoryAlloy cluster-wide KV caching technology and high-performance Rust BPE tokenizer on vLLM. We'll close with a live panel featuring Simon Mo (vLLM/Inferact) and the Crusoe team, covering where inference is heading and what's next for open source AI.

​We'll end the evening with networking, drinks, and bites with the builders scaling the future of AI.

​Schedule:

• ​5:00pm - Doors open

• ​5:30pm - 7:00pm - Talks & Panel Q&A

• ​7:00pm - 9:00pm - Networking Reception

​Note: A government-issued ID is required for building access and check-in for this event.