AI-Powered Python Tutor

• 

  Sign in

• 

  Sign up

• 

  Dashboard

• 

  Sidebar collapsed

• 

  Sidebar expanded

• 

  Curriculum overview

• 

  Topic explanation

• 

  Coding challenge question

• 

  Multiple choice question

Inspiration

In Dr. Tyagi's CS181 lecture the founder of JungleAI gave a presentation on prompt engineering.

What it does

Teaches programming using dynamic, AI generated curriculum. The user is tested using multiple choice and coding challenge questions, where the GPT-4o context analysis beta api and Python interpreter integration analyzes the response, providing effective feedback.

How I built it

React + TypeScript, Firebase for db management and auth, OpenAI api for content generation, Material UI React and React-Bootstrap for UI stuff.

Challenges I ran into

Prompt engineering in practice for the first time was difficult, and making an effective, future proof database schema in a short time.

Accomplishments that we're proud of

The feedback and content generation has been made extremely reliable due to good prompt engineering.

What we learned

Prompt engineering in practice + time management building a fullstack app.

What's next for HowdyHack AI Powered Python Tutor

Support for more languages with simple json configurations, and teacher management so it could be used in classrooms. Response caching and realtime streaming to reduce latency.

Built With

• firebase
• openai
• react
• typescript