Hacker News
Enables AI agents to perform hybrid searches over Hacker News stories, combining keyword and semantic embedding techniques.
소개
Provides a serverless search service for Hacker News, designed to be consumed by AI agents and MCP-enabled clients. It leverages a sophisticated ML pipeline for data ingestion, enrichment, and indexing, offering hybrid search capabilities that combine keyword matching with semantic embeddings. The service features weighted ranking based on Hacker News vote scores and recency, and includes a custom LLM-powered evaluation framework to ensure continuous improvement in search quality.
주요 기능
- ETL pipeline to fetch, enrich, and index Hacker News stories
- 0 GitHub stars
- Custom LLM-powered evaluation framework for iterative search quality improvement
- Hybrid search combining keyword matching and semantic embeddings for high recall
- Serverless deployment on Google Cloud Run for scalability and ease of use
- Weighted ranking based on Hacker News vote score and story recency
사용 사례
- Empowering AI agents with advanced search capabilities over Hacker News content
- Integrating Hacker News search functionality into MCP-enabled client applications
- Developing and testing AI agents that require current and historical Hacker News data
Sponsored