Tertio — AI Wellness Coach
A daily AI companion that guides wellness through personalized, LLM-driven check-ins — designed, built, and shipped end-to-end from zero to the Play Store.
Most wellness apps give you static content — the same daily tips that feel generic after day three.
Founders building in this space needed proof that AI can create genuinely personalised guidance at scale, without needing a human coach on the other end. Users wanted accountability — but on their own terms, not a rigid schedule.
The task: design and ship a mobile app where an LLM serves as an intelligent, empathetic coach that learns from daily check-ins and adapts its response to each user's context, history, and tone.
Research & Scoping
Interviewed 8 potential users. Identified core pain: accountability without rigidity. Defined MVP scope: daily check-in → LLM analysis → personalised response. Mapped the prompt architecture before writing a line of code.
Design & Architecture
Designed the full UI in Flutter — check-in flow, history view, streak tracker. Set up Riverpod state management and Supabase for user data. Iterated on the LLM prompts until responses felt coach-like, not robotic.
Build & Ship
6-week build from first commit to Play Store submission. Clean architecture means v2 features can be added without rewrites. App live with real users. Zero revision disputes. Same process I apply to every founder MVP.