How we turned a citrus disease model into a working MVP
For PhytopathologIA, we transformed a trained PyTorch model into a browser-based demo with image upload, inference, and public access through a budget-conscious delivery stack.
Context
Investor-ready MVP under strict infrastructure constraints.
Stack
PyTorch, lightweight API, web UI, Google Colab, PyNgrok.
Outcome
A live demo stakeholders could test immediately.
Why it matters
- Shows how we scope systems around business timing, not generic best practices.
- Highlights our ability to bridge research artifacts and usable product interfaces.
- Demonstrates the kind of lean delivery that helps early-stage teams validate faster.