Knowledge Catalog and agent-ready business data
Give AI agents accurate, well-described facts about your business
Quick answer: Knowledge Catalog is Google Cloud’s metadata platform, formerly called Dataplex Universal Catalog and renamed in April 2026. It discovers and organises data assets and adds business context, such as descriptions, glossaries, data quality and lineage, so that AI agents can find trusted, well-described data. The same principle applies to any business: AI agents are only as accurate as the facts you give them.
Why it matters
AI agents give wrong answers when the underlying data is vague, duplicated or undocumented. A catalogue of well-described data, with clear business terms and known quality, is what lets an agent answer correctly and lets you check where an answer came from.
Google Cloud’s Knowledge Catalog does this for data held in services such as BigQuery, Spanner, Cloud SQL and Looker. Businesses that are not on Google Cloud still need the same foundation: one accurate, current source of facts about products, prices, policies and opening hours.
If you use Google Cloud
- Metadata review: what is described, what is not, and what an AI agent would struggle to interpret.
- Business glossary: your business terms defined once and linked to the data that represents them.
- Descriptions and ownership: clear descriptions, contacts and quality information on the data assets that matter.
- Access-aware design: so agents only see data they are permitted to see.
- Connecting agents: including through the MCP tools Google provides for Knowledge Catalog.
If you do not use Google Cloud
- A single source of business facts: products, services, prices, policies and contact details, kept accurate in one place.
- Published in machine-readable form: structured data on your website, an
llms.txtguide and, where it helps, an MCP server or A2A Agent Card. - A knowledge base for your own assistant: a curated, reviewable set of questions and answers, the way our own assistant works from the catalogue we publish at /knowledge/catalogue.json.
Sources: Google Cloud: introducing the Knowledge Catalog · Google Cloud documentation: Knowledge Catalog overview. These standards are still evolving, so we check the current specification before we build.
Pricing
Pricing: From £1,500. Every project is scoped and quoted individually after a free call, so the final price depends on your systems and what you need.
Get a quote →Frequently asked questions
What is Google Cloud Knowledge Catalog?
Knowledge Catalog is Google Cloud's metadata management platform, formerly Dataplex Universal Catalog. It discovers data assets, organises them, and adds business context such as descriptions, glossaries, data quality and lineage so that people and AI agents can find and trust data.
Is Knowledge Catalog the same as Dataplex?
It is the renamed Dataplex Universal Catalog. Google renamed it in April 2026, and the API, command-line and permission names did not change.
Do I need Google Cloud to make my business data agent-ready?
No. The principle is the same on any platform: accurate, structured, well-described facts about your business that agents can read. We can build that foundation with or without Google Cloud.
How does this relate to MCP and A2A?
Knowledge Catalog and good business data are what an agent reads. MCP connects an assistant to that data and your tools, and A2A lets agents talk to each other. Accurate data comes first, which is why we start there.
How much does it cost?
From £1,500. Every project is scoped and quoted individually after a free call, so the final price depends on your systems and what you need. Get a quote through our contact page.
Talk to us
Get a quote → Or run the free agent-readiness scan
Or email contact@heimdell-tech-ai.co.uk
Registered in England & Wales, company no. 16478408 · ICO ZC079121