How it works
Four steps: write a brief, approve the spec, we build it, it goes live once every launch check passes.
STEP 1
Brief
Describe your AI chat app in a few lines: who pays, the problem it solves, and the price.
STEP 2
Spec
We turn your brief into a spec: screens, pricing, sign-in and data. Nothing builds until you approve it.
STEP 3
Build
We build from a tested template with sign-in, checkout, and error tracking already wired. API keys never reach the browser.
STEP 4
Go live
Every launch check must pass before the app goes live on its whatasaas.com address. Factory plans can use their own domain.
What ships
Every app is an AI chat app built on Next.js, Better Auth, Polar billing and a Neon database. It sells through your own Polar product. You can add extras like document answers, booking, payments and forms, and they run inside your app. Each app gets a whatasaas.com address, and Factory can use your own domain. Errors are reported to WhatASaaS monitoring; the source download includes Sentry for your own project.
Signing up is free. Writing a brief, building and going live need a paid plan, from $29/mo. Built for web AI apps, not for mobile apps, ERP, custom models, or 99.99% uptime guarantees.
What launch checks are
Before an app goes live, we run up to 17 checks on the build, depending on what the app includes. Every check that applies must pass:
- Spec approved
- Account auth wired (with MFA)
- Payments checkout present
- No secret values in the client bundle
- Sign-in, sign-up, and password reset wired
- Chat requires a signed-in session
- Knowledge search tools present
- Knowledge scoped to this app
- Contact form and email route present
- Declared screens generated
- Source export includes the full scaffold
- Dependency install
- TypeScript check passes
- Production build passes
- Generated tests pass
- Code security scan
- Dependency vulnerability scan
If a check fails, the app stays in preview and we show which check failed.