Ptacek: voice-chat agents turn a plain-English feature request into reviewed code
nptacek · x · 2026-10-16
Thomas Ptacek describes how his team runs multi-human, multi-agent software engineering on a shared non-commodity codebase, pushing back on the idea that AI coding discourse only covers solo vibe coding or lab-scale compute:
- Anyone describes a feature in plain English; agents in a voice chat autonomously map the feature surface and find the best implementation approach.
- The initial design is approved in-conversation; an asynchronous secondary review runs deep checks and regression tests, returning go / no-go / needs-feedback.
- A memory backend and surrounding conversation context anchor the pipeline, and context naturally tightens once it's set up.
He admits orchestration was painful to build, but it drastically reduces touch points.
Related event: Researcher demos voice-directed multi-agent coding workflow(2 posts)→
More from coding & agent
- VibeZero: one-shot an entire vibe coding platform on Cloudflare Void — cnakazawa · 2026-10-16
- Ant and Tsinghua open-source Realtime-Venus, an OpenAI Realtime API alternative — aigclink · 2026-10-16
- 'I remember when I was the harness for o1 pro' — dev on becoming the API for his model — timigod · 2026-10-16
- JevAgentHarness: a 4B decision model lets cheap agents offload tool choices from the LLM — youjiaxuan · 2026-10-16
- Coding agents keep opening 3 stacked PRs per chat, and merging them is a nightmare — holdenmatt · 2026-10-16
- Google-led paper uses sequential auctions to organize decentralized LLM agent workflows — _onionesque · 2026-10-16