A ten-workcell one-person product factory built around OpenAI Dots
A video lays out a full setup around Dots: 10 workcells, 50 reusable jobs and 150 task slots, linking everything from demand signals to retention.
OpenAI Dots is f*cking insane for running a one-person product factory... one founder. three products. one connected loop here's the setup I'd build around it: 10 workcells 50 reusable jobs 150 task slots across the portfolio 1. signal radar → recurring customer pain + sourced briefs 2. opportunity lab → positioning, pricing bets and proof to gather 3. demand engine → pilots, outreach drafts and buyer requirements 4. product blueprint → tight scope, UX, data and acceptance criteria 5. build foundry → working software, previews and tested PRs 6. release control → rollout plans, rollback plans and health checks 7. launch studio → demos, docs, clips and distribution drafts 8. activation desk → onboarding to the first real customer win 9. retention loop → bug reports, scoped fixes and regression proof 10. portfolio intelligence → usage, costs and feedback for the next bet the magic is in the connections: buyer objections rewrite the spec verified code feeds the demo and docs onboarding friction turns into product tasks support tickets become reproducible fixes real usage and costs decide what gets built next each product keeps its own trail: sources → decisions → scope → code → checks → feedback Dots runs the coordination and passes the right context to background tasks Work/Codex environments do the building the 150 slots are reusable templates. run only what you need parallelize what's independent, wait on what isn't, and start with one product in active build you keep the calls that move the system: validation, scope, releases, public posts, outreach and spend save this, then build your solo company with Dots ⭣
This post quotes another post: x.com/mika_systems/status/2107101689878630885