Personal product. I built it because no existing app matched the way I learn
WeSpeak
A progressive web app for remembering the English words you meet every day, with spaced repetition and AI-generated examples.
- Role
- Creator. Product, SQL schema, front end, AI pipeline, deployment
- Period
- June 2026 – present
- Status
- Live and open to sign-ups
- Stack
- React
- Vite
- TypeScript
- Supabase (Auth + PostgreSQL)
- Node / Express
- GitHub Actions
- Vercel
Links
trywespeak.com
The problem
No vocabulary app I tried matched the way I learn: save the words I actually run into during the day, then see them again right before I forget them.
What I built
- Spaced repetition with the SM-2 algorithm, so each word comes back at the right moment.
- LLM-generated content, with examples, audio and context, enriched by a scheduled GitHub Actions job.
- Supabase for authentication and PostgreSQL, on a SQL schema I designed.
- An optional Node/Express backend for server-side AI calls.
- Architecture documented as an interactive HTML map of the codebase.
- An exploration of retrieval-augmented generation (RAG).
My role
I designed and built it end to end: product decisions, the data model, the interface, the AI enrichment pipeline and the deployment.
What’s next
- Verb-tense practice.
- iOS and Android apps built with Capacitor.