After 6 months of development, I'm launching Echo Notes, a tool that transforms audio recordings into organized notes with automatically extracted action items.

The problem I was solving for myself was losing important details in meetings. I'd find myself frantically typing notes while trying to stay present in conversations, ultimately doing both poorly.

Technical implementation: - Built with Next.js and DaisyUI for the frontend - Custom audio visualization using Canvas API with real-time frequency analysis - Backend transcription pipeline using OpenAI's Whisper API - Post-processing with a custom GPT prompt that: 1. Extracts action items and to-dos 2. Identifies people mentioned 3. Generates contextual titles 4. Formats everything into clean markdown

The most challenging part was optimizing the audio processing pipeline to handle varying file formats while keeping the UI responsive.

I'd love feedback from the HN community on the user experience and suggestions for additional features. Currently offering 50% off for early adopters.

Try it at: https://echonotes.fr