VocApp
A mobile-first, API-backed language-learning app built end-to-end for AnzaKen, with gamified memorisation, translation and text-to-speech.
AnzaKen needed a production-ready, API-backed language-learning product built end to end, with secure authentication, validated data and a reliable path to deployment, while adopting AI-assisted, spec-driven engineering practices to keep pace without sacrificing quality.
Owned the product architecture across the full stack: a feature-based Next.js/React frontend and a layered Node.js/Express (TypeScript) backend of routes, controllers, services and repositories, backed by PostgreSQL and Redis. Implemented JWT authentication, input validation and secure API access, and integrated Google Cloud Translation and text-to-speech services for the app's core learning features. Used Claude Code, Codex and Copilot within spec-driven, agentic workflows for implementation, testing and code review, with Docker and GitHub Actions for CI/CD and Sentry for observability.
VocApp is live and in active development, with an architecture and CI/CD pipeline that supports continued feature delivery through the same AI-assisted, spec-driven workflow.