Gorse

An open-source AI-powered recommendation engine. Provides a web dashboard for managing items, users, and recommendation strategies with real-time system monitoring.

Getting Started

This template deploys Gorse in all-in-one playground mode, which bundles all components (master, worker, server) and an embedded database into a single container. It automatically imports sample data from GitRec for demonstration.

What You Can Do After Deployment

1. Visit your domain — the Gorse dashboard opens at the root URL
2. Explore the dashboard — view system status, recommendation pipeline, and data overview
3. Manage items and users — add, edit, or import items and users via the dashboard or REST API
4. Configure recommendation strategies — tune collaborative filtering, content-based, and popularity-based algorithms
5. Integrate via REST API — fetch personalized recommendations for your application at /api/recommend/:user_id
6. Monitor system metrics — track recommendation quality, training progress, and cache status

Key Features

• Collaborative filtering (matrix factorization, nearest neighbors)
• Content-based recommendation via embeddings
• LLM-powered rankers for advanced personalization
• Real-time recommendation serving via REST API
• Web dashboard for pipeline management and monitoring
• Auto-tuning of model parameters
• Multi-source data import (CSV, JSON, API)

API Examples

# Get recommendations for a user
curl https://YOUR_DOMAIN/api/recommend/user_id

# Insert feedback
curl -X POST https://YOUR_DOMAIN/api/feedback \
  -H "Content-Type: application/json" \
  -d '[{"FeedbackType":"star","UserId":"bob","ItemId":"item_1"}]'

License

Apache-2.0 — GitHub