Knowledge Agents

What is a Knowledge Agent?

Agent that augments Vector Search and LLM capabilities with enterprise data, enabling accurate, contextrich responses by retrieving relevant information from internal knowledge bases, documents, or web-sources.

Core Capabilities

• Contextual retrieval from unstructured sources and files

• Grounded responses traceable to enterprise-approved sources

• Supports unauthenticated web sources for broader knowledge retrieval

• Supports file system sources for knowledge retrieval

• Creates context based suggestions to ask the Agent

• Web crawling support for dynamic data sources

• Metadata detection for PDF documents

Create a Knowledge Agent

1. Add Data Sources:

   Create the knowledge sources for your agent to learn from by adding one or more unstructured data sources, such as file uploads, Microsoft SharePoint document repositories, or web content, including unauthenticated web pages. See Data Sources for instructions on adding website and file system data sources that serve as the agent’s knowledge base.

2. Automated Data Processing:

   Each data source you configure is automatically processed and progresses through the following stages.

   1. Crawling: Retrieve data from configured sources.

   2. Parsing: Analyze unstructured data and transform into structured format.

   3. Storing: Write the structured data from parsing stage to the database.

   4. Chunking: Divide the text data into chunks for direct embedding in the next stage.

   5. Embedding: Convert collected data into vector representations.

   6. Ingestion: Store processed data in the vector database.

   Note: Data processing happens as soon as a data source is configured and are not part of the Knowledge Agent creation.

3. Select Data Sources:

   Once the configured data is fully processed and ingested, it becomes available for selection as a data source for the Knowledge Agent. You can choose any number of data sources, including a mix of web, SharePoint, and file system sources.

4. Create Knowledge Base:

   Assign a name and description to the knowledge agent. Make the description as detailed as possible because it will be used for many downstream purposes.

5. Publish the Agent and Start Conversing:

   Publish the agent to make it available. Chat with the knowledge agent and receive answers based on the ingested data.