01> HOW IT SHIPPED
// THE CHALLENGE01
Indian eating does not survive a calorie tracker designed around Western portions and packaged food, and a nutrition assistant that invents a dish is worse than no assistant at all.
// THE APPROACH02
One Flutter codebase for both stores over a self-managed FastAPI and Postgres backend, with pgvector and ParadeDB holding an Indian food corpus in the same database. Every AI answer is retrieved from that corpus and cited back to it rather than generated free-hand.
// THE OUTCOME03
Questionnaire, plan and daily tracking ship as one loop, with photo-based logging on top. Guidance stays grounded in real dishes and real portions, which is the only version of this product worth trusting.
case --info idyora
> client · IDYORA
> sector · Health & Wellness
> stack · Flutter · FastAPI · PostgreSQL · pgvector · RAG
> result · iOS + ANDROID · ONE CODEBASE · LIVE
> case/idyoraSHIPPED ▮
Want a system like this?
Tell us what you're building. We'll respond with a scoped plan, not a sales pitch.