EcoSphere

Campus sustainability issues get reported through scattered channels and nobody can see what happened next. This is a product prototype for what a single reporting, tracking and reporting-back surface would look like.

• Twelve App Router routes covering dashboard, reporting, issue tracking with per-ticket detail, sustainability metrics, admin centre, waste scanner, predictions, rewards, notifications, profile and settings.
• Global client state through a provider holding issues, eco points, notifications and toasts over seeded mock data.
• Forms built with React Hook Form and Zod validation, charts with Recharts, transitions with Framer Motion.
• A deterministic simulation engine in lib/ai-engine.ts: a seeded xorshift generator hashed from the input, so the same issue or image always produces the same analysis, with replies assembled from phrase pools rather than returned by a model.
• Photo uploads read client-side through FileReader, so nothing is transmitted anywhere.

No machine learning model

The AI-labelled features are produced by a deterministic offline simulation engine in lib/ai-engine.ts. There is no trained model, no inference and no external API call anywhere in this project.

Seeded mock data held in client-side React state. There is no backend, no database and no persistence: a refresh resets the application.

Runs as a complete frontend prototype. All twelve routes work against seeded state and the simulation engine returns consistent analyses for the same input. There is no authentication, backend, persistence or machine learning model.

This is a product engineering project, not an ML one. Its AI-styled features are simulated by design, so there are no model metrics to report and none are implied.

• Made the simulation engine deterministic, seeded from a hash of the input, so the same report or image always yields the same analysis and demos stay reproducible.
• Documented the simulated AI prominently in the repository README rather than presenting generated text as model output.