Noted

Icon for the app

Inspiration:

Capturing and understanding information from lectures, meetings, or conversations is often overwhelming. People forget key points or lose context because they are trying to listen, think, and take notes at the same time. We wanted to build a tool that removes that burden and makes it easy to revisit what matters.

What it does:

Noted turns your smart glasses into a simple but powerful memory assistant. Right now, it lets you upload audio recordings for transcription and ask questions about the content using the Gemini API. This creates a clear, searchable text record of conversations or lectures that you can refer back to anytime.

How we built it:

Backend (FastAPI): Handles audio uploads and transcription, as well as question answering powered by the Gemini API. Frontend (React + Vite): Provides a simple web interface where users can upload recordings and interact with the AI. Mentra Glasses Client: Acts as the capture device that records and sends audio to the backend for processing.

Challenges we ran into:

A teammate left 3 hours in. Getting accurate and fast transcription from real-world audio recordings. Managing audio file uploads and connecting them reliably to the backend. Designing an interface that feels simple but still useful at this early stage. Accomplishments that we're proud of Built a working end-to-end system that can upload audio, transcribe it, and let users query the content using Gemini. Created a solid foundation for future features like visual capture, bookmarking, and automatic summaries.

What we learned:

Even simple capabilities like transcription and question answering can unlock powerful use cases. Reliable pipelines for audio handling and processing are essential before layering on more complex features. Iterating quickly on core functionality helps shape a clearer roadmap for the next stages.

What's next for Noted:

Adding visual capture and OCR from MentraOS smart glasses. Building a searchable timeline that connects audio and visual context. Implementing highlights, summaries, and flashcard generation. Expanding the interface into a richer review and study platform.