Show HN: Local-first video intelligence for Mac – find by text, visual, or audio

Category: ai-ml

Tags: video-search, local-first, semantic-search, ai, docker, self-hosted

Score: 7.8/10 (Innovation: 7, Technical: 8, Documentation: 8, Utility: 8)

Edit Mind (Orcah) is a local-first video intelligence platform that indexes video content using AI-powered transcription, frame analysis, and multi-model embeddings. It enables natural language search across spoken words, visual objects, and scenes, running fully locally via Docker for privacy. The project combines multiple ML models and vector search to create a personal video knowledge base, with a commercial desktop app planned for easier deployment.

Target audience: video editors, content creators, privacy-conscious users, self-hosters, developers

Repository: https://orcah.app/ · TypeScript · NOASSERTION · 1815 stars

View on Hacker News