IDP @OneTutor: Prototype the future of higher education

Shape an early-stage startup, work on cutting-edge AI, and make university life better for thousands of students.

👋🏼 About Us

We’re OneTutor – an AI-powered learning platform built by students, for students. At TUM, you might know us as TUMTutor. Our intelligent AI tutor helps students make sense of their lectures and is already in use at 20+ universities with over 10,000 learners.

Now, we’re inviting you to join our journey, through an Interdisciplinary Project (IDP) at the intersection of AI, Education, and Startups.

• 👋🏼 About Us
• 🪄 What we do
• 🛠️ What You’ll Work On
• 📅 Project Timeline
• 🎯 What you bring
• 🌟 Why OneTutor
• ✅ How to Apply

🪄 What we do

We strive to make one-on-one, personalized tutoring accessible to all students by leveraging responsible AI.

OneTutor delivers personalized, AI-driven tutoring for university courses via:

• RAG-powered Chat: Instantly answers student questions with context-aware, reliable responses sourced directly from lecture materials using Retrieval-Augmented Generation.
• Dynamic Quizzes: (Semi-)automated generation of tailored quizzes with instant grading and constructive feedback, helping students actively test and deepen their understanding.
• Learning by Doing, Not Just Watching: We go beyond passive content consumption. OneTutor makes lecture materials interactive, turning static slides into a dynamic learning journey where students engage, respond, and reflect.

Learn more: https://onetutor.ai/

🛠️ What You’ll Work On

Depending on your interests, you can work on one of the following projects and shape it with your own ideas:

1. Data-Driven Quality Improvements

Build an evaluation pipeline that uses real user data to assess the quality of our AI answers. You’ll analyze usage patterns, identify performance bottlenecks, and define measurable quality metrics. Based on your findings, you’ll develop concrete strategies for improving system reliability and quality.

2. AI & Content Retrieval - Making Videos Searchable

The core challenge: integrating lecture videos (not just transcripts) into a searchable AI system. You’ll develop strategies for handling multi-modal content, videos and slides into a unified retrieval pipeline. Your work will focus on extending our AI backend and optimizing RAG pipelines to enable seamless search, retrieval, and interaction with mixed content formats.

3. User Experience

Analyze the existing user flows for both students and lecturers. Identify UX/UI bottlenecks, prioritize key pain points, and develop concrete improvement plans. You’ll design, prototype, and implement adjustments that make the platform more intuitive and enjoyable.

📅 Project Timeline

• Wk 1: Onboarding & planning your project
• Wk 2–4: Build, test, break (building the first MVP)
• Wk 5: Drafting the MVP into a scalable solution
• Wk 6-10: Implementation phase
• Wk 11: Reflection & testing
• Wk 12: Final presentation & wrap-up

🎯 What you bring

• You’re a student who actively uses AI tools to enhance your learning.
• Startup reflexes – bias for action, high autonomy, low ego.
• Strong TypeScript skills, ideally across the full stack (with tools like: React, NestJS, LangchainJS).
• Experience or curiosity around LLMs and RAG

🌟 Why OneTutor

• Build something you’d actually use as a student
• Get hands-on startup experience
• High autonomy with regular check-ins and supervision as needed
• Join a passionate team at our cozy office next to Theresienwiese

✅ How to Apply

1. Send us a short CV or your LinkedIn profile.
2. Attach your Transcript of Records
3. Write one paragraph on your preferred project and why OneTutor + you = 🚀.
1. As an AI startup, we appreciate it if your paragraph reflects your own thoughts, not generic AI output 😉

👉🏼 Email everything to join@onetutor.ai with subject “IDP – [Your Name]”