• ACCURACY
  • CONTEXT
  • RELIABILITY
  • HUMAN

How to ground your AI responses in AWS data

Connect your AI to your Amazon Bedrock knowledge bases. Once linked, your agent uses your existing documents to answer questions accurately.

Ask AI about this page

Short answer

How can I make my AI use my AWS documents for answers?

You link your Amazon Bedrock knowledge bases to your AI client. The agent then queries those specific data sources to find the right information before it speaks. You get answers backed by your own files.

The results.

Where your answers come from.

The AI uses your Bedrock knowledge bases to turn raw documents into reliable answers.

ACCURACY

Fact-based answers

The agent uses your Bedrock data to find specific context. This stops the AI from making things up when you ask about your private files.

CONTEXT

Deep document search

Your AI scans through your existing knowledge bases. It finds the exact snippets needed to satisfy your specific technical or business queries.

RELIABILITY

Grounded responses

Every answer is tied to the information found in your AWS storage. You get a direct line between the question and your actual documentation.

HUMAN

Manual review needed

The AI provides the text based on your data, but you decide if the interpretation is correct. You remain the final judge of the output.

The workflow

What your AI does when the data arrives.

The agent moves from searching your files to generating a response based on what it finds.

  1. Search your data

    The agent looks through your Amazon Bedrock knowledge bases to find relevant information related to your prompt.

    retrieve
  2. Generate answers

    The agent takes the retrieved document snippets and uses them to write a coherent response to your question.

    retrieve_and_generate
  3. Verify context

    The agent checks the retrieved segments to ensure they actually address the user's specific request.

    retrieve
  4. Finalize output

    The agent combines the retrieved facts into a single, direct answer for you to read.

    retrieve_and_generate

Try it

Copy these to start.

Use these prompts to test your connection immediately.

Starting points

These are basic ways to start. Swap out the bracketed text with your specific document names or topics.

Amazon Bedrock Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_VAqgKgPAMaCVADpiaKW1YIUR9Y3n0l8xHvlUnkn1/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Amazon Bedrock capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "amazon-bedrock-kb-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_VAqgKgPAMaCVADpiaKW1YIUR9Y3n0l8xHvlUnkn1/mcp"
    }
  }
}
Copy into chat04
  • Based on the [Project Alpha] documentation in my knowledge base, what are the deployment requirements?

  • Summarize the latest updates found in my [Q3 Engineering] AWS data.

  • Search my Bedrock knowledge base for any mention of [API Authentication] protocols.

  • Using the information in my [Employee Handbook] data, how do I request time off?

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Amazon Bedrock KB once, then ask.

Just follow the link to connect your AWS account. Your credentials stay encrypted, and you can start querying your data immediately.

Connect Amazon Bedrock KB to your AI

FAQ

How the grounding works

  • 01

    Can the AI change my documents in AWS?

    No. The tools only allow the AI to read and retrieve information from your knowledge bases. It cannot edit, delete, or upload files to AWS.

  • 02

    Does the AI use my data to train its general model?

    No. The connection is used only to provide context for your specific queries within your session.

  • 03

    What happens if the information isn't in my knowledge base?

    If the agent cannot find the answer in your Bedrock data, it will tell you it doesn't know rather than guessing.

  • 04

    Can I use this with any AI client?

    Yes, as long as you are using an MCP-compatible client like Claude, Cursor, or Windsurf.

  • 05

    Does this work with S3 buckets directly?

    No. This connector specifically queries Amazon Bedrock Knowledge Bases. Your data must be indexed in a Bedrock KB first.

  • More questions about Amazon Bedrock KB? See everything the Amazon Bedrock KB Connector can do

Connect Amazon Bedrock to Claude, Cursor, ChatGPT & more