AmazonBedrockKnowledgeBaseRetriever

Retrieves documents from an Amazon Bedrock Managed Knowledge Base.

Most common position in a pipeline	1. Before a ChatPromptBuilder in a RAG pipeline 2. The last component in the semantic search pipeline
Mandatory init variables	knowledge_base_id: The ID of the Amazon Bedrock Knowledge Base. Falls back to the AWS_KNOWLEDGE_BASE_ID env var.
Optional init variables	aws_access_key_id: AWS access key ID. Can be set with AWS_ACCESS_KEY_ID env var. aws_secret_access_key: AWS secret access key. Can be set with AWS_SECRET_ACCESS_KEY env var. aws_session_token: AWS session token. Can be set with AWS_SESSION_TOKEN env var. aws_region_name: AWS region name. Can be set with AWS_DEFAULT_REGION env var. aws_profile_name: AWS profile name. Can be set with AWS_PROFILE env var. number_of_results: Maximum number of results to return. Defaults to 5. use_agentic_retrieval: If True, tries the Agentic Retrieve API before falling back to the standard Retrieve API. Defaults to the USE_AGENTIC_RETRIEVAL env var, or True.
Mandatory run variables	query: A string
Optional run variables	top_k: Maximum number of results to return. Overrides number_of_results if provided.
Output variables	documents: A list of Documents
API reference	Amazon Bedrock
GitHub link	https://github.com/deepset-ai/haystack-core-integrations/tree/main/integrations/amazon_bedrock/
Package name	amazon-bedrock-haystack

Overview

AmazonBedrockKnowledgeBaseRetriever retrieves Documents from an Amazon Bedrock Managed Knowledge Base. Unlike most other Retrievers, it doesn't need a Haystack Document Store or an Embedder: indexing and embedding are handled entirely by AWS, and the component only needs a text query to search the Knowledge Base.

By default, the Retriever tries the Agentic Retrieve API first and falls back to the standard Retrieve API if agentic retrieval isn't available for the configured Knowledge Base. You can control this behavior with the use_agentic_retrieval init parameter, or the USE_AGENTIC_RETRIEVAL environment variable.

Each returned Document includes a score and metadata about where it came from: source (the S3, web, Confluence, Salesforce, SharePoint, or custom document location of the underlying content), knowledge_base_id, and knowledge_base_type.

This component uses AWS for authentication. You can use the AWS CLI to authenticate through your IAM. For more information on setting up an IAM identity-based policy, see the official documentation.

If the AWS environment is configured correctly, the AWS credentials are not required, as they're loaded automatically from the environment or the AWS configuration file. If the AWS environment is not configured, set aws_access_key_id, aws_secret_access_key, and aws_region_name as environment variables or pass them as Secret arguments.

Installation

Install the Amazon Bedrock integration:

bash

pip install amazon-bedrock-haystack

You also need an existing Amazon Bedrock Knowledge Base with documents already ingested. Set its ID as the AWS_KNOWLEDGE_BASE_ID environment variable, or pass it directly as the knowledge_base_id init parameter.

Usage

On its own

python

from haystack.utils import Secret

from haystack_integrations.components.retrievers.amazon_bedrock import (
    AmazonBedrockKnowledgeBaseRetriever,
)

retriever = AmazonBedrockKnowledgeBaseRetriever(
    knowledge_base_id="ABCDEFGHIJ",
    aws_region_name=Secret.from_token("eu-central-1"),
)

result = retriever.run(query="What are the benefits of managed knowledge bases?")
for doc in result["documents"]:
    print(doc.content)
    print(doc.meta["source"])
    print(doc.score)

In a RAG pipeline

python

from haystack import Pipeline
from haystack.components.builders import ChatPromptBuilder
from haystack.dataclasses import ChatMessage
from haystack.utils import Secret

from haystack_integrations.components.generators.amazon_bedrock import (
    AmazonBedrockChatGenerator,
)
from haystack_integrations.components.retrievers.amazon_bedrock import (
    AmazonBedrockKnowledgeBaseRetriever,
)

template = [
    ChatMessage.from_user(
        """
    Given these documents, answer the question.\nDocuments:
    {% for doc in documents %}
        {{ doc.content }}
    {% endfor %}

    \nQuestion: {{question}}
    \nAnswer:
    """,
    ),
]

rag_pipeline = Pipeline()
rag_pipeline.add_component(
    "retriever",
    AmazonBedrockKnowledgeBaseRetriever(
        knowledge_base_id="ABCDEFGHIJ",
        aws_region_name=Secret.from_token("eu-central-1"),
    ),
)
rag_pipeline.add_component(
    "prompt_builder",
    ChatPromptBuilder(template=template, required_variables="*"),
)
rag_pipeline.add_component(
    "llm", AmazonBedrockChatGenerator(model="global.anthropic.claude-sonnet-4-6")
)

rag_pipeline.connect("retriever.documents", "prompt_builder.documents")
rag_pipeline.connect("prompt_builder.prompt", "llm.messages")

question = "What are the benefits of managed knowledge bases?"
result = rag_pipeline.run(
    {
        "retriever": {"query": question},
        "prompt_builder": {"question": question},
    },
)
print(result["llm"]["replies"][0].text)