Stepscape

Apr 2025
Stepscape
Stepscape
Stepscape

Objective

The goal of this project is to allow users to get their 10,000 steps in with things they would be interested in doing, anywhere in the world to avoid repetitive or boring walks.

Tools & Technologies

Python, FastAPI (API backend), Railway (backend hosting), Fly.io (frontend hosting), Streamlit (Frontend UI), OpenAI API (interest parsing), Google API (geocoding and POIs)

Challenge

Stepscape was built as part of the Headstarter program, where we were challenged to come up with and create a full-stack, public-facing project in just one week.

Living in NYC, I've always appreciated how walkable the city is. But even in a place like this, walks can start to feel repetitive and it's not always easy to plan a walk with purpose. That's what inspired Stepscape.

Stepscape lets you enter a starting location: an address, a neighborhood, or just a point of interest. Then you tell it whether you'd like to end up near where you started, and what you'd like to experience along the way. Your interests can be typed in freely, there's no fixed format. An LLM processes your input and the system figures out a route with stops tailored to your prompts, in the order you specify.

Behind the scenes, the app uses a FastAPI backend with a structured, multi-step workflow to interpret your interests, match them to real-world locations, and generate a walking route. It returns a Google Maps URL with personalized stops, all triggered via simple POST requests. The experience works seamlessly on both web and mobile.

This was a deeply rewarding experience. I'd never built something this complex before, involving a deployed backend and microservices. I learned a ton about end-to-end software development and I'm incredibly proud of the final result. Check it out, go for a walk, and tell me what you think!