TaniDoc Nufarm

Overview

TaniDoc is a crop-diagnostic platform built as two applications: TaniDoc for farmers and TaniDoc Expert for plant experts. It combines AI-assisted diagnosis with direct consultation — a farmer can identify a plant problem and talk it through with a specialist.

The challenge

The product had to work for farmers who aren't necessarily technical, in the field, sometimes on patchy connections — while still supporting real-time chat with experts and an AI diagnosis flow. Making something that sophisticated feel simple was the core challenge.

My role

• Founding frontend engineer — owned the entire client side from the ground up, alongside a dedicated backend engineer and a UI/UX designer.
• Mobile: built the Flutter apps for farmers (TaniDoc) and experts (TaniDoc Expert) end to end.
• Internal web: built the React admin/operations side the team runs day to day.
• Owned the user-facing flows and the real-time features linking farmers to experts.

Approach

I kept the farmer's path short — diagnose, then talk to an expert — and used Socket.IO for chat so answers arrive live. GetX keeps mobile state predictable, and Firebase handles the realtime/auth plumbing.

What I built

• AI crop diagnosis for identifying plant problems from user input.
• Real-time consultation and chat between farmers and experts over Socket.IO.
• Supporting features: auth, e-commerce, and a directory of nearby kiosks.

Under the hood

• Mobile: Flutter with GetX, a Socket.IO client, and Firebase.
• Web: React + TypeScript, React Hook Form + Yup, React Bootstrap, Redux.

Tools

Figma · VS Code · Postman · Git · Trello · VPS · PM2 · Google Play Console