XINGO
AI role-play practice for spoken interpreting. Two AI participants who can't understand each
other, one human interpreter, an instant examiner-style score. Built for NAATI CCL candidates
and working community, medical and legal interpreters.
Stack: Next.js 16 (App Router) · React 19 · Tailwind v4 · Convex (database + backend) · Clerk (auth) · Stripe (payments) · OpenAI Realtime + Responses · GA4.
Docs
| |
|---|
| docs/product.md | What XINGO is, who it's for, the experience, pricing, roadmap (PRD v2) |
| docs/architecture.md | System design, data model, practice engine, metering, billing, environments |
| docs/decisions.md | Decision log (why things are the way they are) |
| docs/design-system.md | Visual language, tokens, components, copy voice |
| docs/analytics.md | GA4 events and how to read the funnel |
| docs/findings-2026-10.md | Production data review, security fixes, team feedback log |
| docs/known-issues.md | Open risks and follow-ups |
| docs/runbooks/stripe-setup.md | Step-by-step payments setup |
| docs/runbooks/release-v4.md | Cutover to the production deployment, admin access, rollback |
Run locally
pnpm install
npx convex dev # starts/attaches your dev Convex deployment and watches convex/
pnpm dev # Next.js on http://localhost:3000
.env.local needs NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY, CLERK_SECRET_KEY (Clerk Development instance keys), NEXT_PUBLIC_CONVEX_URL and CONVEX_DEPLOYMENT (written by npx convex dev). Server secrets (OpenAI, Stripe, Clerk issuer) are Convex environment variables, set with npx convex env set — see architecture §8.
First time on an empty deployment:
npx convex run seed:seedBaseData
npx convex run users:setRole '{"email":"[email protected]","role":"platform_admin"}' # after signing in once
Checks
pnpm typecheck
pnpm lint
pnpm test # Convex business rules (convex-test)
pnpm build
CI runs typecheck, lint and tests on every pull request (.github/workflows/ci.yml).
