AI Neighbourhood Finder
RAG-powered Edmonton neighbourhood explorer with semantic search over civic data and an interactive map.
Choosing where to live in a city you just moved to means cross-referencing parks, transit, schools and community data that live in a dozen unconnected open-data portals. This tool puts them behind one conversational interface.
Neighbourhood profiles are embedded into a Chroma vector store; a retrieval-augmented pipeline answers free-text questions — 'quiet, walkable, close to an LRT stop' — by grounding the model's answer in the retrieved civic data rather than letting it guess. Results render on an interactive map, and clicking any neighbourhood starts a conversation about it.
Built for MM 802 (Multimedia Communications) and deployed on Cybera Rapid Access Cloud.
Highlights
- RAG over Edmonton open civic data with ChromaDB embeddings.
- Conversational advisor grounded in retrieved neighbourhood profiles.
- Interactive Google Maps interface with click-to-query neighbourhoods.
- Deployed to an OpenStack instance on Cybera Rapid Access Cloud.