Hawker Opportunity Score Platform

R3izorr/sc2006-proj
repo

Full-stack geospatial platform scoring business opportunity for Singapore's hawker centres. Built for SC2006 with Vite/React, FastAPI, PostGIS-style data and an AI assistant layer.

TypeScript React Vite Tailwind CSS Python FastAPI Leaflet PostgreSQL #full-stack #school #sc2006 #geospatial
TypeScript 1 updated Nov 2025 school

Group project for NTU SC2006 (Software Engineering). The product answers one question: “If I wanted to open a stall, which hawker centre gives me the best shot?”

What it does

  • Pulls hawker-centre data and public datasets, joins them to geospatial boundaries, and scores each centre by demand, competition, and supporting amenities.
  • Displays results on a Leaflet map and in a sortable, filterable dashboard.
  • Lets a user draw a custom service area and recompute the score.
  • Has a small AI assistant that explains why a score was given and suggests what to look at next.

Stack split

  • Frontend (/client): Vite + React + TypeScript, Tailwind, Leaflet for maps.
  • Backend (/server): FastAPI with typed routers, SQLAlchemy, scoring services, geospatial helpers.
  • Data: seeded from public datasets; schema is geo-aware so score recomputation stays fast.

What I worked on

  • Scoring pipeline and API design.
  • Map interaction layer (hover states, per-centre overlays, isolines).
  • The assistant’s “explain this score” flow on top of the scoring service.