Project Cafecito / Beans
Beans is our aggregation and search service for news and blogs. It supports semantic query search, practical filtering, and clean JSON output. Think of it as your content firehose with better manners.
Fresh content from over 7,000 news and blog publishers, updated daily. We handle the full aggregation pipeline so you get clean, structured data without any crawling or maintenance overhead.
Natural language vector-based semantic search with tunable accuracy. Go beyond keywords — ask questions in plain text and pull deeply relevant articles from thousands of sources. Ideal for powering AI agents, RAG systems, and intelligent automation workflows.
Rich, multi-dimensional filtering by categories, named entities, geographic regions, content types, sources, and time ranges. Designed for precise control while remaining simple to use in real AI and analytics workflows.
Every article is automatically scored and ranked based on real social engagement signals. Surface what’s actually gaining traction instead of just the most recent content.
Automatic extraction of named entities, sentiment, topics, and regions for every article. Ready for analytics dashboards, AI agents, and downstream pipelines — no extra processing layers required.
Automatically links related articles across thousands of publishers. Track how stories spread from niche sources to mainstream outlets and observe narrative shifts in real time.