Predictive Analytics Platform & ML Service

A full-stack predictive analytics platform that turns raw data into ML-powered forecasts, delivered as a subscription SaaS. I built the entire stack: a Python and FastAPI backend serving the machine-learning models, a React and TypeScript dashboard for end users, secure JWT and OAuth2 authentication, and Stripe-powered billing.

The client needed to productise a machine-learning model into a real, paying SaaS — not just a notebook. That meant a secure API around the model, user accounts and authentication, subscription billing, and a clean dashboard that non-technical users could actually use. It had to be production-ready and safe to launch.

A Python / FastAPI backend exposing the predictive model through a fast, typed REST API

A React / TypeScript dashboard for running predictions and visualising results

JWT and OAuth2 authentication for secure, multi-user access

Stripe integration for subscriptions and recurring billing

A production-ready architecture with clear separation between the ML service and the application layer

Python, FastAPI, React, TypeScript, and Stripe, with JWT/OAuth2 authentication.

The client went from a standalone machine-learning model to a fully working, monetised SaaS — with secure authentication, subscription billing, and a polished dashboard, all built and shipped as a single full-stack engagement.