NYC Civil Navigator

An AI assistant bridging New Yorkers to real-time open data for SNAP markets, Wi-Fi hubs, and local services in plain English

Comment

Inspiration

Navigating municipal services, local events, and public datasets in NYC can be overwhelming for everyday residents. While NYC Open Data provides incredible public datasets, raw data isn't easily accessible to non-technical users. We were inspired to build NYC Civic Navigator to bridge this gap using generative AI—allowing anyone to query city services, SNAP markets, Wi-Fi spots, and library programs in plain, natural English.

What it does

NYC Civic Navigator is an intuitive AI assistant designed for all New Yorkers:

  • Natural Language Querying: Users can ask everyday questions like "Where can I find SNAP-eligible markets?" or "How do I access free public Wi-Fi?"
  • Pre-set Query Buttons: Instant prompt shortcuts for non-technical users who want quick access to popular city resources without typing.
  • Grounding in Open Data: Connects AI responses to live NYC Open Data endpoints, ensuring accurate responses accompanied by direct source citations.

How we built it

  • Intelligence: Powered by Google AI Studio API (gemini-1.5-flash) for fast, natural language reasoning and prompt grounding.
  • Backend: Built with Python for quick prototyping, API orchestration, and data handling.
  • Frontend: Lightweight and accessible web UI designed for speed and ease of use.
  • Data Source: NYC Open Data (Socrata SODA API) for live public datasets.

Challenges we ran into

  • Balancing raw API data payloads with context limits to keep LLM responses fast and cost-effective.
  • Crafting system prompts that translate dense, municipal data fields into simple, plain-English answers accessible to everyday residents.

Accomplishments that we're proud of

  • Successfully combining "vibe coding" with public civic data to build a functional prototype within a weekend.
  • Making open-source city data genuinely useful and accessible for non-technical New Yorkers.

What we learned

  • How to effectively leverage Google AI Studio to quickly ground generative model responses in dynamic local datasets.
  • The power of rapid prototyping (vibe coding) to quickly turn a civic idea into a working tool.

What's next for NYC Civic Navigator

  • Integrating geographic distance/mapping to rank resources by proximity to the user's current borough or neighborhood.
  • Expanding multi-language support so non-English speaking communities across NYC can access city services seamlessly.

Built With

Share this project:

Updates

Submission history