Skip to main content
Knowledge bases allow agents to ground their responses in your documents. This is essential for building agents that can answer questions about your specific data.

​
What is a knowledge base?

A knowledge base stores and retrieves documents:
  • Documents are chunked and embedded
  • Vector database stores embeddings for similarity search
  • Agents can search the knowledge base with agentic RAG

​
Quick example

from agno.agent import Agent
from agno.knowledge import Knowledge
from agno.vectordb.pgvector import PgVector

# Create knowledge base
knowledge = Knowledge(
    path="docs/",
    vector_db=PgVector(
        table_name="agno_docs",
        db_url="postgresql://localhost/agno"
    )
)

# Load documents
knowledge.load(recreate=False)

# Create agent with knowledge
agent = Agent(
    knowledge=knowledge,
    search_knowledge=True,
    add_knowledge_to_context=True
)

response = agent.run("What is Agno?")

​
Using ChromaDB (development)

from agno.knowledge import Knowledge
from agno.vectordb.chroma import ChromaDb

knowledge = Knowledge(
    path="docs/",
    vector_db=ChromaDb(
        collection="docs",
        path="./chroma_db"
    )
)

knowledge.load()
See cookbook examples for complete patterns.

Build docs developers (and LLMs) love

Get started for freeTalk to us