EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Sep 1·5 min read
Keenable SELECT: an agent that searches the web in SQL
TL;DR
Keenable SELECT is an innovative agent that enables SQL-based web searches, allowing users to extract structured data from diverse online sources. It provides detailed reports on research movements in AI labs since 2025, enhancing data accessibility for researchers.
✦ Why It Matters
Start using Keenable SELECT to track AI researcher movements and identify potential collaboration opportunities today.
Key Takeaways
- Utilize Keenable SELECT for efficient web data extraction.
- Leverage SQL queries to filter and structure research data.
- Implement the dual-agent system for automated report generation.
- Access previous query results easily for ongoing research.
How It Works
Keenable SELECT operates by executing SQL queries on live web data through a read-only DuckDB SELECT statement. It employs web and semantic operators to gather and filter data before processing it in DuckDB.
The system can handle multiple queries simultaneously, merging results and removing duplicates, which allows for efficient data extraction with minimal LLM calls.