How I built NotesGPT – a full-stack AI voice note taking app

NotesGPT is a full-stack AI voice note-taking app developed by Hassan El Mghari, which has garnered significant attention with 35,000 visitors and 7,000 users shortly after its launch. The app allows users to record voice notes, transcribe them using Whisper, and extract action items via Mixtral, all while being fully open source and mobile-responsive. The technological foundation of NotesGPT includes Convex for the database and cloud functions, Next.js for the application framework, and Together.ai for inference and embeddings. Authentication is facilitated by Clerk, while Tailwind CSS is used for styling. The app's architecture is designed for efficiency, with a dashboard for managing notes, vector search capabilities for semantic searching, and server-side rendering to keep data updated. The backend setup includes defining schemas for data handling and utilizing Replicate to host Whisper for transcriptions. Additionally, the app uses Together.ai to generate summaries and action items from transcriptions, offering a comprehensive note management solution.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.