After months of building in private, we are opening Kognita to early access teams.
What is Kognita?
Kognita is a managed knowledge base API. You push documents in; we handle chunking, embedding, and indexing. You query with natural language; we return the most relevant content using hybrid BM25 + vector search.
The goal is to make the retrieval layer of any RAG pipeline a solved problem, so you can focus on what your application actually does.
What is available today
- Knowledge base management: create, update, and delete knowledge bases via API or dashboard
- Content ingestion: push plain text, HTML, or Markdown; we handle chunking and embedding
- Hybrid search: BM25 + vector similarity with a configurable
alphablend parameter - MCP server: expose your knowledge base as a tool any MCP-compatible agent can call
- Embedding model selection: choose from multiple embedding models per knowledge base
- API key management: scoped keys with per-key usage tracking
What is coming next
- Semantic chunking: automatic topic-aware splitting at ingestion time
- Embedding migration: zero-downtime migration between embedding models
- Webhooks: get notified when ingestion completes
- PDF and DOCX ingestion: native support without pre-processing
How to get started
Sign up at kognita.io and follow the quickstart guide. Early access teams get direct access to the team for onboarding support.
We are building this for teams who want reliable retrieval without managing vector infrastructure. If that is you, we would love to hear what you are building.