Edit Mind lets you search your videos by content, not just filenames. Recognize faces, transcribe speech, detect objects, and explore your library with natural language search. All locally and securely.

Perfect for creators, editors, and researchers who need smarter video management.

⚠️ Development Status: Edit Mind is currently in active development and not yet production-ready. Expect incomplete features and occasional bugs. We welcome contributors to help us reach v1.0!

Click to watch a walkthrough of Edit Mind's core features.

• Search videos by spoken words, objects, faces, and events.
• Runs fully locally, respecting privacy.
• Works on desktop and web.
• Uses AI for rich metadata extraction and semantic search.

• Video Indexing and Processing: A background service watches for new video files and queues them for AI-powered analysis.
• AI-Powered Video Analysis: Extracts metadata like face recognition, transcription, object & text detection, scene analysis, and more.
• Vector-Based Semantic Search: Powerful natural language search capabilities on video content using ChromaDB and Google Gemini.
• Dual Interfaces: Access your video library through a native Desktop App (Electron) or a Web App (Docker).

This project is structured as a pnpm monorepo, separating concerns into distinct applications and shared packages.

• apps/desktop: The native Electron application, providing a rich user experience.
• apps/web: A full-stack web application for browser-based access.
• apps/background-jobs: The core backend service managing video processing, AI analysis orchestration, and job queues. (Used for the Docker setup)

• packages/prisma: Database schema and migration management.
• packages/shared: (Under refactoring) Contains utilities, types, and services shared across applications.
• packages/ui: A shared UI component library (Under construction to share components between web and the desktop application).

• python/: Contains Python scripts for various AI-powered video analysis plugins, transcription, face matching and face reindexing. It's communicating via WebSockets.

Edit Mind uses Docker Compose to run everything in containers.

• Docker Desktop installed and running.
• That's it! Everything else runs in containers.

git clone https://github.com/iliashad/edit-mind
cd edit-mind

Important: Before proceeding, configure Docker to access your media folder.

macOS/Windows:

1. Open Docker Desktop
2. Go to Settings → Resources → File Sharing
3. Add the path where your videos are stored (e.g., /Users/yourusername/Videos)
4. Click Apply & Restart

Linux: File sharing is typically enabled by default.

Edit Mind uses a two-file environment configuration:

• .env - Your personal configuration (required)
• .env.system - System defaults (required)

Copy the example file and customize it:

cp .env.example .env

Edit the .env file and configure these critical settings:

# 1. SET YOUR VIDEO FOLDER PATH (REQUIRED)
# Must match the path you added to Docker File Sharing
HOST_MEDIA_PATH="/Users/yourusername/Videos"

# 2. CHOOSE AI MODEL (Pick one option)
# Option A: Use Gemini API (easier, requires API key)
USE_LOCAL_MODEL="false"
GEMINI_API_KEY="your-gemini-api-key-from-google-ai-studio"

# Option B: Use Local Model (more private, requires model download)
# USE_LOCAL_MODEL="true"
# SEARCH_AI_MODEL="/app/models/path/to/.gguf"
# The AI model should be downloaded and saved it to models folder in the project root dir

# 3. GENERATE SECURITY KEYS (REQUIRED)
# Generate with: openssl rand -base64 32
ENCRYPTION_KEY="your-random-32-char-base64-key"
# Generate with: openssl rand -hex 32
SESSION_SECRET="your-random-session-secret"

Quick Key Generation:

# Generate ENCRYPTION_KEY
openssl rand -base64 32

# Generate SESSION_SECRET
openssl rand -hex 32

cp .env.system.example .env.system

Start all services with a single command:

docker compose up

First-time startup will take 5-10+ minutes as Docker:

• Builds all container images
• Downloads AI models and dependencies
• Initializes databases

Once all services are running (look for "ready" messages in logs):

• Web App: http://localhost:3745
• BullMQ Dashboard: [http://localhost:4000/(http://localhost:4000) (Job queue monitoring) if you have NODE_ENV set to development

1. Navigate to the web app at http://localhost:3745
2. Login using admin@example.com and password is admin
3. Navigate to the web app at http://localhost:3745/app/settings
4. Click "Add Folder"
5. Select a folder from your HOST_MEDIA_PATH location
6. The background job service will automatically start processing your videos and will be start watching for new video file events
7. Monitor progress in the BullMQ dashboard at http://localhost:4000

Problem: "Empty section between colons" error

# Solution: Ensure .env is copied to docker directory
cp .env docker/.env
# Verify HOST_MEDIA_PATH is set
grep HOST_MEDIA_PATH .env

Problem: Services won't start

# Check Docker is running
docker --version
docker compose version

# View detailed logs
docker compose -f docker/docker-compose.yml logs

# Check service status
docker compose -f docker/docker-compose.yml ps

Problem: Cannot access video files

# Verify HOST_MEDIA_PATH is correct
ls -la /your/video/path

# Check Docker file sharing includes this path
# Docker Desktop → Settings → Resources → File Sharing

# Restart Docker after adding paths

For the native Electron desktop experience, see apps/desktop/README.md. Note: The desktop app requires local Node.js and pnpm installation.

.
├── apps/                 # Individual applications (desktop, web, background-jobs)
│   ├── background-jobs/  # Node.js service for AI analysis orchestration & job queue
│   ├── desktop/          # Electron desktop application
│   └── web/              # Full-stack web application
├── packages/             # Shared libraries and packages
│   ├── prisma/           # Prisma schema, migrations, and database utilities
│   ├── shared/           # Cross-application constants, types, and utilities
│   └── ui/               # Reusable UI components
├── python/               # Core Python AI/ML analysis services and plugins
├── docker/               # Dockerfiles and docker-compose configurations
└── ...                   # Other configuration files (pnpm-workspace.yaml, .env.example, etc.)

For detailed instructions on each application, refer to their respective README.md files:

• apps/desktop/README.md
• apps/web/README.md
• apps/background-jobs/README.md

I would like to thank the community of/r/selfhost for their amazing support and feedback (https://www.reddit.com/r/selfhosted/comments/1ogis3j/i_built_a_selfhosted_alternative_to_googles_video/)

We welcome contributions of all kinds! Please read CONTRIBUTING.md for details on our code of conduct and the process for submitting pull requests.

This project is licensed under the MIT License - see the LICENSE.md file for details.