Eagle Eye MapChat
Capstone GIS assistant with voice, radio-style input, and AI-powered map actions
Radio Workflow Demo
Live Diarization
Technologies Used
Project Overview
Eagle Eye MapChat was my capstone project: a Flutter-based field GIS assistant designed for teams like wildfire response and search-and-rescue. The app lets users talk to a map in natural language, routes those requests through OpenAI Assistants, and turns them into map queries or edits across mapping backends like Leaflet and CalTopo. My work focused on the radio and chat experience, helping make the system feel more practical for real field communication instead of just a demo chatbot.
Challenges
The hardest part was making a voice-first mapping workflow feel reliable in the field. That meant handling speech capture, message timing, transcript cleanup, persistent chat history, and map-aware assistant responses in a way that still felt fast and usable under pressure.
Solution
I worked on the chat and radio pipeline by wiring together push-to-talk style speech capture, transcription flow, Firebase-backed messaging, and assistant-driven map actions. The system supports both voice and text input, stores conversation history, and gives users a more natural way to interact with complex GIS data.
Impact & Results
The project became a strong capstone demonstration of how conversational AI can lower the barrier to GIS tools for field teams, especially when hands-free or low-friction communication matters.