Knowledge Base
Connect any data source to your AI. PDF, websites, databases, Markdown, and CSV — auto-chunked and vectorized for accurate, context-aware AI responses.
- PDF Documents — Import and search PDF files with AI-powered chunking and vectorization.
- Website Crawling — Crawl and ingest websites in real time to keep your AI context current.
- Markdown Files — Import Markdown docs, wikis, and structured content into your Knowledge Base.
- SQL Database — Connect SQL databases directly for AI-powered queries and insights.
- CSV & Spreadsheets — Import CSV files and spreadsheets into searchable AI context.
- Auto-Chunking — Intelligent document splitting for optimal AI retrieval accuracy.
- Vectorization Engine — High-performance vector embeddings for accurate semantic retrieval.
- Semantic Search — Find information by meaning, not just keywords, across all sources.
- Real-Time Sync — Keep your Knowledge Base current with automatic data synchronization.
FAQ
What data sources can I connect to a Knowledge Base?
You can ingest PDFs, websites, markdown files, CSV spreadsheets, SQL databases, and more. The platform automatically chunks and vectorizes your data for accurate AI responses.
How does Knowledge Base search work?
Knowledge Bases use semantic vector search — your data is converted to embeddings and stored in a vector database. When an AI agent needs information, it performs a similarity search to find the most relevant content.
How many Knowledge Bases can I create?
The Starter plan includes 1 Knowledge Base (1M tokens). Plus plan offers 15 Knowledge Bases (200K each). Pro plan gives unlimited Knowledge Bases at 1M+ scale each.