Inspiration

People often overlook how everyday environmental conditions, such as heat and air quality, can affect their health and well-being, especially those with existing climate-sensitive health conditions. While climate data is increasingly available, it remains difficult to translate into actionable decisions.

What it does

CliMap NYC is an AI-assisted website that combines a user’s calendar, self-reported personal context, location, and environmental forecasts to generate evidence-informed recommendations. Through interaction with an intelligent conversational planning agent, real-time environmental data is translated into simple, personalized, actionable guidance on when, where, and what to do.

How we built it

Real-time EPA Air Quality Index data, Forty Guard heat-map data, and COPD-related health data inform the baseline model used for environmental risk scoring. A protected LLM powers the planning agent, which takes desensitized personal context and uses the baseline risk scores to generate personalized recommendations. The interface was prototyped in Lovable and implemented with Claude Code.

What's next for CliMap NYC

With more support, we would develop the prototype into a functional pilot by refining and validating the personalization and risk-scoring framework. We would expand the model to incorporate additional environmental risks, such as pollen and flooding, and develop the AI activity-planning system to learn users’ preferences over time while keeping objective environmental and public-health evidence at the foundation of its recommendations.