GooseChase

Inspiration

Runners today rely on apps like Strava that track data well but don’t actually coach. Most feedback is surface-level and generic. We wanted to build something that feels like a real running coach, one that understands your data, adapts to you, and helps you improve over time.

What it does

GooseChase is an AI-powered running coach that analyzes your recent runs and gives personalized, data-driven feedback. Instead of static plans or generic summaries, it looks at patterns like pace trends and effort to provide actionable insights, suggest adjustments, and guide your training like a real coach would.

How we built it

We built GooseChase using a frontend interface (React + Tailwind) connected to an LLM that processes structured run data. We simulate or integrate activity data (like distance, pace, and heart rate) and feed it into the model along with a coaching framework inspired by established training methods. The system then generates contextual, personalized coaching responses.

Challenges we ran into

One of the biggest challenges was making the AI feel like a real coach instead of just a chatbot. We had to carefully design prompts so responses were grounded in actual data and training principles. Another challenge was structuring raw activity data in a way the model could interpret meaningfully.

Accomplishments that we're proud of

We’re proud that GooseChase goes beyond basic tracking and actually delivers coaching-style insights. The system can reference specific run data, adapt its feedback, and ask follow-up questions, making it feel interactive and personalized rather than static.

What we learned

We learned how important context and structure are when working with LLMs. Simply giving data isn’t enough, how you frame it determines the quality of output. We also learned how to balance technical implementation with product design to create something that feels intuitive and useful for the user.

What’s next for GooseChase

Next, we’re planning to add features like a Pace Prophet that predicts your PR based on your training, a race readiness score that shows how prepared you are as you progress, and injury risk estimation to help runners avoid overtraining and stay consistent. Long-term, we see GooseChase becoming a fully autonomous AI coach that guides runners through their entire training journey.

Built With

• claude
• next.js
• shadcn
• strava
• supabase
• tailwind
• vercel