GreenPath is a personal travel-time memory for recurring journeys. I built it to make the hidden part of a daily drive visible: time repeatedly lost at traffic lights and while searching for parking.
After a trip, GreenPath lets me review candidate stops and confirm which were traffic lights. It also records parking-search time for familiar destinations, then combines those histories into a more realistic hidden-delay estimate, with morning or evening context when enough observations exist.
The app is intentionally lightweight and private: it runs as a no-build web app, stores history locally in the browser, and includes a deterministic sample. In the sample run, a Morning Commute to Riverside Plaza records three confirmed traffic-light stops totaling 4:20, followed by a 5:10 search for parking. GreenPath combines both histories to show 9:30 of expected hidden time for the journey. I used Codex with GPT-5.6 to turn the idea into scoped, testable milestones, design a safer post-trip classification flow, preserve local-storage compatibility, and build repeatable simulation and validation.
Log in or sign up for Devpost to join the conversation.