The goal
Build an AI-powered learning product where people can practise IELTS speaking and visa interviews.
My role
As CTO, I led product engineering, technical architecture, AI infrastructure, deployment, and reliability. This was a team product; my work spanned the application and the systems behind it.
What we built
I built and shipped production features across Flutter and web, LLM inference, speech-to-text, text-to-speech, and real-time AI interactions.
The work included Supabase and PostgreSQL, Python services, Docker, AWS, and cloud deployment workflows.
Engineering decisions
My responsibilities included product architecture, analytics, infrastructure debugging, model and API integrations, and AI usage-cost optimisation. The challenge was making those parts work together in a usable product.
The output
Voyage AI is available on Google Play. The published listing provides an external record of the product and its interface.
What I learned
Retrospective to come. I’m keeping the project record factual until I add specific decisions, tradeoffs, and lessons from the work.
Role and contribution: supplied resume. Product and screenshot: public Google Play listing, inspected September 2026.