Studr - AI Notetaker

Inspiration 🚀

As a student and builder, I’ve struggled with capturing lectures, rewriting notes, and staying organized. Traditional notetaking apps are static — they don’t help you learn. With the rise of AI, I wanted to build something that transforms lectures and materials into searchable, summarized, and interactive study aids.

The vision for Studr – AI Notetaker came from this question:

What if students could instantly turn lectures into searchable, summarized, and interactive study material?

What it does ✨

Studr – AI Notetaker helps students learn faster by turning raw content into structured knowledge.

• 🎤 Record lectures & voice notes
  
• ✍️ Get instant, accurate transcriptions
  
• 🤖 Summarize notes with AI
  
• 📄 Upload PDFs & documents and ask questions
  
• 🎥 Summarize YouTube videos (via proxy to avoid blocking)
  
• 🔍 Search across all study materials instantly
  
• ☁️ Sync across devices with cloud storage
  

How I built it ⚙️

• Frontend: Flutter (iOS + Android)
  
• Backend: Flask (Python)
  
• AI/ML: AssemblyAI for transcription, OpenAI GPT for summarization and Q&A
  
• Vector DB: LangChain + Qdrant for embeddings and semantic search
  
• Background tasks: Redis + Upstash for job queues
  
• Database: Supabase
  
• Payments: RevenueCat
  
• Hosting: Hetzner VPS (self-hosted API + server management I learned during this build)
  

Challenges I ran into 🧩

• Achieving transcription accuracy in noisy classroom environments
  
• Learning how to set up Redis & Upstash for background task processing
  
• Scaling semantic search across long lecture transcripts
  
• Deploying APIs on Hetzner and handling server ops for the first time
  
• Navigating App Store/Play Store privacy + subscription compliance
  
• Balancing simplicity vs power in the student UI
  

Accomplishments that I’m proud of 🏆

• Built a full end-to-end AI notetaking app in hackathon time
  
• Learned to manage my own server infrastructure (Hetzner VPS)
  
• Integrated multiple complex APIs (AssemblyAI, OpenAI, RevenueCat) into one seamless flow
  
• Created a unique feature: YouTube summarization with proxy bypass
  
• Designed an app that actually solves a real student problem
  

What I learned 📚

I learned more during this build than in months of tutorials:

• How to combine multiple AI services into one cohesive product
  
• Setting up and scaling background queues with Redis + Upstash
  
• Using LangChain + Qdrant for embeddings and semantic search
  
• Hosting and maintaining a production API on Hetzner VPS
  
• Why clear privacy practices are critical for education apps
  
• That the hardest part isn’t coding — it’s building something students will trust and actually use
  

What’s next for Studr – AI Notetaker 🔮

• Add multi-language transcription for international students
  
• Expand AI with quizzes and flashcard and other cool features
  
• Improve scalability to handle thousands of students
  
• Launch to the Directories/Launchpad for wider adoption
  
• Build a strong student feedback loop to guide new features
  

Built With

• flutter
• hetzner
• python
• supabase