Gergely Orosz · 7 min read · AI & Product
Building Codex with Tibo Sottiaux
Read the original on newsletter.pragmaticengineer.com →
On the Pragmatic Engineer Podcast, Tibo Sottiaux discusses how Codex was built and evolved at OpenAI, including the Rust-based CLI, multi-provider model support, and its integration into the software lifecycle. The conversation also surfaces product-management lessons from Tibo's career and the trade-offs behind building AI-powered tooling.
Key Takeaways
- Shock-cancellation underscored the importance of product-market fit, strong user feedback loops, and not overreading PM optimism.
- Google experimented with a 'ChatGPT-like' internal project early on, but it never shipped, illustrating how early ideas don’t guarantee market-ready products.
- OpenAI appealed to him because ChatGPT was built by a small team (~20 engineers), highlighting the value of tight Research–Product collaboration.
- Codex uses Rust to prioritize performance, security, and scale across millions of cloud instances, even though AI models were stronger in Python/TypeScript at the time.
Excerpt
Stream the latest episode Listen and watch now on YouTube , Apple , and Spotify . See the episode transcript at the top of this page, and timestamps for the episode at the bottom. Brought to You by • turbopuffer – a vector and full-text search engine built on object storage. It’s fast, cheap, and extremely scalable. The teams building the smartest AI products out there — Anthropic, Cognition, Notion, Harvey — all run on turbopuffer. Check it out • Antithesis – use AI agents to work on critical systems without worrying about correctness. Antithesis runs your complete system in…