Sherpa AI Learning

Inspiration

Learning is not one-size-fits-all, yet most online learning platforms still deliver the same content to every learner. We wanted to build an AI tutor that understands how each person learns, instead of simply answering questions. Inspired by the role of a mountain sherpa, our goal was to create an AI that guides learners through their own learning journey—adapting explanations, adjusting difficulty, and helping them progress step by step rather than giving instant answers.

What it does

Sherpa is a personalized AI learning platform that adapts to each learner's goals, interests, current knowledge, and learning style. Learners can choose any topic they want to study. Sherpa explains concepts, generates practice questions, evaluates responses, and adjusts the learning path based on performance. The platform also includes: Learning Map to visualize progress and future learning paths. Mistake Book to automatically collect concepts that require more review. Learning Resources to recommend deeper reading and related materials. Instead of being just another chatbot, Sherpa provides a complete adaptive learning experience.

How we built it

We designed the product from the user experience first, starting with the learning workflow before implementing the technical architecture. The frontend was built with Next.js, React, and TypeScript, while Supabase handled authentication and database management. We integrated the OpenAI API to power personalized explanations, question generation, and adaptive feedback. Throughout development, we also collaborated with ChatGPT and Codex to accelerate product design, architecture planning, debugging, and implementation.

Challenges we ran into

One of our biggest challenges was designing an adaptive learning flow instead of simply connecting an LLM to answer questions. We also needed to organize learner progress into meaningful structures such as the Learning Map and Mistake Book while keeping the user experience simple and intuitive. Another challenge was integrating AI into a real product for the first time, including API configuration, prompt design, and maintaining a smooth learning experience.

Accomplishments that we're proud of

-Built a working personalized AI learning platform. -Created a complete adaptive learning workflow instead of a simple chatbot. -Designed visual learning features like the Learning Map and Mistake Book. -Successfully integrated OpenAI into a real educational application. -Completed the project within the limited Build Week timeline.

What we learned

This project taught us that building an AI product is much more than calling an API. Great AI experiences require thoughtful product design, user experience, prompt engineering, and continuous iteration. We also learned how AI tools such as ChatGPT and Codex can significantly improve collaboration and development efficiency.

What's next for Sherpa AI Learning

Our next goal is to make Sherpa even more personalized. We plan to support long-term learner memory, richer knowledge graphs, personalized study plans, spaced repetition, and multimodal learning with voice and images. We also hope to introduce deeper progress analytics so Sherpa can become a true long-term learning companion rather than just an AI tutor.

Built With

• adaptive
• ai
• api
• chatgpt
• codex
• css
• education
• learning
• next.js
• openai
• personalized
• react
• supabase
• tailwind
• typescript