For OpenAI's ChatGPT
Stuck in ChatGPT? Press the dot.
The most-built-on track. The press is for the moments after 'it works'.
When ChatGPT needs a person.
01
When the OpenAI tool-use chain breaks
Tool calls that worked in dev fail in prod. Function-call schemas drift. The press surfaces an engineer who's traced these before.
02
When the prompt's right but the latency isn't
Streaming, structured output, model selection across the OpenAI lineup — a Relay engineer profiles the call path and trims the round-trip.
03
When the build needs to ship
From notebook to production: env vars, secrets, observability, the deploy. The engineer takes the press through the cutover.
Common questions
- Do you support GPT-5 / GPT-4o?
- Yes — every model in the OpenAI lineup, including the realtime API and the Assistants API.
- What about ChatGPT plugins / GPTs?
- Yes. Building a custom GPT or wiring a ChatGPT plugin to a real backend is one of the most common press patterns we see.
- Can the engineer help with Azure OpenAI?
- Yes. Common in regulated industries; the engineer covers the routing, the auth, and the latency-sensitive bits.
Related reading
- Field notes · 9 min readAnatomy of the handoff: what we learned shipping the first 1,000 sessionsA median match in seconds isn’t a marketing number; it’s a system constraint we engineered toward and missed by 16 seconds in the first month.
- Research · 17 min readWhen does an AI build want a person? A study of 4,200 stuck momentsWe logged every press in the first quarter of the private beta. Four patterns explain almost everything.
- Essay · 11 min readThe irreducibly human moment in softwareAn argument for why software engineering doesn’t shrink in the age of AI, it sharpens, and moves to a new place in the build.