Cassandra Vector Database

Use Cassandra as a vector database for your Knowledge Base.

Setup

Install cassandra packages

1uv pip install cassandra-driver

Run cassandra

1docker run -d \
2  --name cassandra-db \
3  -p 9042:9042 \
4  cassandra:latest

Example

1from kern.agent import Agent
2from kern.knowledge.knowledge import Knowledge
3from kern.vectordb.cassandra import Cassandra
4from kern.knowledge.embedder.mistral import MistralEmbedder
5from kern.models.mistral import MistralChat
6from cassandra.cluster import Cluster
7
8# (Optional) Set up your Cassandra DB
9
10cluster = Cluster()
11
12session = cluster.connect()
13session.execute(
14    """
15    CREATE KEYSPACE IF NOT EXISTS testkeyspace
16    WITH REPLICATION = { 'class' : 'SimpleStrategy', 'replication_factor' : 1 }
17    """
18)
19
20knowledge_base = Knowledge(
21    vector_db=Cassandra(table_name="recipes", keyspace="testkeyspace", session=session, embedder=MistralEmbedder()),
22)
23
24knowledge_base.insert(
25    url="https://kern-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf"
26)
27
28agent = Agent(
29    model=MistralChat(provider="mistral-large-latest", api_key=os.getenv("MISTRAL_API_KEY")),
30    knowledge=knowledge_base,
31)
32
33agent.print_response(
34    "What are the health benefits of Khao Niew Dam Piek Maphrao Awn?", markdown=True, show_full_reasoning=True
35)

Async Support ⚡

Cassandra also supports asynchronous operations, enabling concurrency and leading to better performance.

1import asyncio
2
3from kern.agent import Agent
4from kern.knowledge.embedder.mistral import MistralEmbedder
5from kern.knowledge.knowledge import Knowledge
6from kern.models.mistral import MistralChat
7from kern.vectordb.cassandra import Cassandra
8
9try:
10    from cassandra.cluster import Cluster  # type: ignore
11except (ImportError, ModuleNotFoundError):
12    raise ImportError(
13        "Could not import cassandra-driver python package.Please install it with pip install cassandra-driver."
14    )
15
16cluster = Cluster()
17
18session = cluster.connect()
19session.execute(
20    """
21    CREATE KEYSPACE IF NOT EXISTS testkeyspace
22    WITH REPLICATION = { 'class' : 'SimpleStrategy', 'replication_factor' : 1 }
23    """
24)
25
26knowledge_base = Knowledge(
27    vector_db=Cassandra(
28        table_name="recipes",
29        keyspace="testkeyspace",
30        session=session,
31        embedder=MistralEmbedder(),
32    ),
33)
34
35agent = Agent(
36    model=MistralChat(),
37    knowledge=knowledge_base,
38)
39
40if __name__ == "__main__":
41    asyncio.run(knowledge_base.ainsert(
42        url="https://kern-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf",
43        )
44    )
45
46    # Create and use the agent
47    asyncio.run(
48        agent.aprint_response(
49            "What are the health benefits of Khao Niew Dam Piek Maphrao Awn?",
50            markdown=True,
51        )
52    )

Tip

Use aload() and aprint_response() methods with asyncio.run() for non-blocking operations in high-throughput applications.

Cassandra Params

Parameter	Type	Default	Description
`table_name`	`str`	`None`	Name of the table to store vectors and metadata in Cassandra
`keyspace`	`str`	`None`	Keyspace name in Cassandra where the table will be created
`embedder`	`Optional[Embedder]`	`OpenAIEmbedder()`	Embedder instance to generate embeddings
`session`	`CassandraSession`	`None`	Active Cassandra session object for database operations