Inspiration

Tennis coaches with 20+ students spend 30 to 40 hours a week on unpaid lesson planning. Tools like Calendly and CoachNow each solve one problem in isolation. No one has built the all-in-one. We wanted to change that.

What it does

A web app that generates a fully personalized lesson plan in under 30 seconds. The app tracks student history, logged conditions, and coach preferences to produce structured and actionable plans.

How we built it

We used GoVector for data flow and state management across the app's layers, and Claude via the Anthropic API as the intelligence engine. Claude takes in each student's full context and returns a structured lesson plan with a built-in reasoning card so coaches always understand why the plan was generated that way.

Challenges we ran into

Our two biggest hurdles were keeping app features consistent across the app without redundancy, and integrating AI in a way that felt native rather than bolted on. Getting Claude to return reliable, structured output took significant prompt engineering and iteration.

Accomplishments that we're proud of

We shipped real coach and client sign up flows, a live dashboard with active client management, session debriefs, and AI generated lesson plans all working together. Integrating Claude into a structured, multi layer app and getting it to return consistent, usable output was a challenge we are genuinely proud of solving under pressure.

What we learned

AI is most powerful when it augments expertise rather than replacing it. Strong data architecture is what makes AI actually useful. The model is only as good as the context you give it.

What's next for CourtSide (Team Mild)

The foundation is built. Next we want to expand the calendar integration so post session debriefs trigger automatically, deepen the intelligence layer so the app surfaces patterns across a coach's full client roster over time, and refine the AI output so lesson plans get smarter the longer a coach uses the platform. Long term, CourtSide has the potential to become the all in one operating system for private tennis coaches.

Built With

  • ai
  • autonomous
  • backend
  • frontend
  • govector
  • python
  • vector
Share this project:

Updates