Scaling AI Infrastructure

​Building and running AI applications at scale takes more than a good model — it takes infrastructure you can actually reason about, version, and ship with confidence. Join Pinecone and Pulumi for an evening of talks on how modern AI systems are built and operated under the hood.

​We'll dig into the real engineering behind production AI: managing the infrastructure that powers vector search and RAG, and using Infrastructure as Code to deliver services reliably and repeatably.

​Talks

• ​How Pinecone Builds Infrastructure — A behind-the-scenes look at how Pinecone leans on Infrastructure as Code to deliver its services. (Speaker: Joerg Schad, VP Engineering)

• ​I Built an AI Running Coach With Pulumi — Every year I race my friend Malcolm at an ultramarathon. Every year he beats me. So I built a system to fix that. This talk walks through Momentum, the AI running coach I built and use daily. It lives in Slack, tracks my training, and adapts its coaching based on how my week is actually going. Not a ChatGPT wrapper. An event-driven system that merges messy real-world data to build context the model actually needs. (Speaker: Adam Gordon Bell, Community Engineer)

​Expect practical, demo-driven sessions, time for Q&A, and plenty of space to hang out with the speakers and other folks working on AI infra afterward.

​Who should come Engineers, platform and infra folks, and anyone curious about what it actually takes to run AI services in production.