What is Data Commons?
Data Commons is the current Blueprint navigation label for public knowledge bases. A knowledge base is a collection of documents or data sources prepared for search and retrieval. It can provide grounded context to prompts, Blueprints, and chatbots instead of relying only on a model’s general knowledge.Browse public knowledge bases
- Open Public Libraries and select Data Commons, or go directly to iblueprint.ai/knowledge-bases.
- Search with Search names, descriptions, categories, and tags….
- Use the Newest sort control to change ordering.
Review a knowledge-base card
Cards show the name, visibility, description, category, tags, document count, and total size. Select View Details to inspect the resource before using it. Treat the public description as a summary and verify the underlying sources, scope, freshness, and access requirements.Use a public knowledge base
Select Use this KB when the resource meets your needs. Blueprint may require sign-in and will add or copy the resource into the workspace flow supported by your account. Review the resulting connection before sending real queries.Connect knowledge to a chatbot
In a chatbot’s Knowledge configuration, link the knowledge base or one of its document sets. Review the available retrieval controls, such as enabled state, priority, maximum chunks, similarity threshold, and retrieval method, then test the chatbot with questions that should and should not be answered from the source.Good practice
- Confirm the source owner, permitted use, update cadence, and document scope.
- Prefer specific document sets when a chatbot should use only part of a knowledge base.
- Test citations, unsupported questions, stale content, and conflicting sources.
- Keep confidential, personal, and regulated information out of public resources.
- Review permissions, retention, and downstream sharing before connecting a knowledge base to a chatbot or workflow.
A knowledge base improves grounding, but it does not make an answer automatically correct. Validate the source and preserve a human review path for consequential use.