​🦄 ai that works

​A weekly conversation about how we can all get the most juice out of todays models with @hellovai & @dexhorthy

​/ai-that-works/ai-that-works

​On #19 we talked about workflows for interruptible agents. One of the most popular segments was on how to combine a small model for a voice agent, with a smarter+slower instruction-following model to observe the conversation and inject steering into the smaller model's context window.

​On this week's episode, we'll dive deep into the architecture of how you can build supervisor-threaded systems, and ship real working code you can pull down and use to build your own high-quality voice agents.

​Pre-reading

​To prevent repeating the basics, we recommend you come in having already understanding some of the tooling we will be using:

• ​Discord

• ​Cursor or VS Code

• ​Programming languages

  • ​Application Logic: Python or Typescript or Go

  • ​Prompting: BAML (recommend video)

​Meet the Speaker 🧑‍💻

​​​Meet Vaibhav Gupta, one of the creators of BAML and YC alum. He spent 10 years in AI performance optimization at places like Google, Microsoft, and D. E. Shaw. He loves diving deep and chatting about anything related to Gen AI and Computer Vision! 

Meet Dex Horthy, founder at HumanLayer and coiner of the term Context Engineering. He spent 10+ years building devops tools at Replicated, Sprout Social and JPL. DevOps junkie turned AI Engineer.