A Knowledge Source connects business information to AI-powered experiences in Yeeflow. AI Agents and Copilots can use configured knowledge to provide answers grounded in relevant application content.
With AI-assisted creation, you can describe the source information and its intended use, then review the proposed configuration before creating it.
Before You Start
Open the application where you want to create the Knowledge Source. Make sure you have permission to add application resources and access the information you want to use.
Before starting, identify:
The existing resource containing the information.
The fields or content relevant to your use case.
Any conditions that should limit the information included.
How an AI Agent or Copilot will use the knowledge.
For example, you might use an existing Document Library containing internal policies or a Data List containing product information.
Creating a Knowledge Source does not create the underlying business information. Prepare the relevant documents or records first.
Step 1: Open New Application Resource
In your application, click Add resource (+) to open New application resource.
In the Create with AI area, select the knowledge resource option. Depending on the interface, this may appear as Knowledge or Knowledge source. Click Show more if it is not immediately visible.
Step 2: Describe Your Knowledge Requirements
Enter a description of the Knowledge Source you want to create.
Identify the existing resource and explain what information AI should use and why.
The resource and field names in this example are illustrative. Replace them with names that exist in your application. Review the plan to confirm which content and settings can be configured for your selected source.
For better results, include the following details:
Detail | What to include |
Knowledge Source name | A clear name, such as Employee Policy Knowledge |
Source resource | The existing resource containing the information |
Intended use | Questions or tasks the knowledge should support |
Relevant content | Fields, documents, or other information AI needs |
Inclusion conditions | Rules that determine which information is included |
Exclusions | Information that should be left out |
Click Start with AI to submit your request.
Step 3: Clarify the Source and Scope
AI may ask questions to resolve missing or ambiguous information.
For example, it may need to clarify:
Which resource to use when similar names exist.
Which fields contain the information you described.
Whether a requested status or category exists.
Which records or documents should be included.
Respond with the actual resource names and business rules from your application.
For example:
Use the HR Policies Document Library in the Employee Services application. The status field is named Policy Status, and the published value is Published.
If the source does not contain a requested field, clarify whether to use another field, omit the condition, or update the source separately.
Step 4: Review the Implementation Plan
AI prepares an Implementation Plan based on your requirements and any clarifications.
Review the proposed Knowledge Source configuration before confirming creation.
Review area | What to check |
Name and purpose | The Knowledge Source is clearly identified |
Source resource | The correct application resource is selected |
Included information | The proposed content is relevant to the intended use |
Conditions | Inclusion and exclusion rules match your requirements |
Additional setup | Any remaining configuration is identified |
If changes are needed, explain them in the conversation.
Review the revised plan before proceeding.
The Knowledge Source is created only after you confirm the Implementation Plan.
Step 5: Confirm and Create the Knowledge Source
When the plan meets your requirements, confirm it in the AI conversation.
AI creates the Knowledge Source based on the confirmed plan. Follow any additional setup prompts shown in the interface.
When creation is complete, open the Knowledge Source using the resource link or entry provided.
Step 6: Check the Configuration and Processing Status
Open the Knowledge Source and check that:
The correct source resource is selected.
Relevant fields and content are included.
Any configured filters match your requirements.
The source contains the expected documents or records.
Review any processing or synchronization status shown in the interface. Complete any required setup and resolve reported issues before relying on the knowledge in an AI experience.
Creating the configuration does not necessarily mean all source content is immediately ready for retrieval. Check the available status information before testing.
Step 7: Connect and Test with an AI Agent or Copilot
After the Knowledge Source is ready, add it to the intended AI Agent or Copilot through its knowledge configuration.
Do not assume that creating a Knowledge Source automatically connects it to every Agent or Copilot.
Test with questions that the source information can answer.
For example:
What information must employees provide when submitting an expense claim?
Check whether the response accurately reflects the source documents. Where references are available, review them to verify the answer.
Also test questions outside the source’s scope to check how the Agent or Copilot handles missing information.
Continue Refining with Ask AI
After creation, use Ask AI to return to the original saved conversation and request further changes.
Review the updated configuration and any processing status, then test the affected questions again.
Tips for Better Results
Start with a Clear Use Case
Explain what users need to ask or accomplish.
For example:
Help employees find answers about leave eligibility and application requirements.