All work

    DietTalk

    A doctor-supervised diet and weight management app, live on the App Store and Google Play.

    Role
    Mobile and full-stack engineer
    Timeline
    June 2025 to present
    Stack
    Flutter, iOS + Android, Web portal, PostgreSQL, Supabase, OCR + LLM extraction
    DietTalk, Mobile and web

    The problem

    DietTalk needed to work for two very different users at once: patients tracking their diet and weight from their phone, and doctors supervising them and sending guidance. That is a mobile app and a clinical web tool, tied to the same patient data, and it had to ship to both app stores.

    What I built

    DietTalk is a doctor-supervised diet and weight management product. Patients use a Flutter mobile app to track diet and weight; doctors use a web portal to supervise and send guidance, with AI insights on top. It is live on the App Store and Google Play.

    I built the mobile app in Flutter for both iOS and Android, the doctor-facing web portal, and the backend on PostgreSQL and Supabase that connects them.

    The hard parts

    One codebase, two stores

    One Flutter codebase had to ship cleanly to both iOS and Android and pass both stores, while the doctor portal stayed tied to the same patient records the app writes to.

    Reading documents instead of re-typing them

    Instead of making people re-type everything, the product reads lab results and captures patient details straight off documents, then uses that to auto-fill the clinical workflow. It removes a pile of manual data entry, and it means the data doctors act on comes from the source. Health data is handled securely throughout.

    Where it is now

    DietTalk is live on the App Store and Google Play, in real use by patients and the doctors supervising them.

    Building something with AI, or stuck on one that keeps breaking? Tell me where it's stuck.

    zeeshan@auxonai.comBook a call