The platform page is switched off for {{product.name}}. A product team can turn it on in site.config.js by setting features.platformPage to true.
THE PLATFORM
{{product.description}}
A DAY WITH {{product.nameUpper}}
An illustrative day with {{product.name}}, built from the workflows it runs today. A steward asks in Teams, the agent checks the live systems and the work lands where the team already tracks it.
“Who owns the finance table, and is it documented?”
A steward asks who is responsible for a table. The agent looks up the data owner and data steward in Collibra, finds the documentation page in Confluence and checks whether the table is registered. One answer, drawn from three systems.
The steward asks for a quality check on a table. The agent reads the quality threshold from the policy in Confluence, measures column completeness in Azure SQL and confirms the table is catalogued in Collibra.
Where a check fails, the agent finds the responsible person and creates a Jira ticket assigned to them, labelled and prioritised by severity. The work lands in the team’s existing backlog, not in a separate tool.
Collibra holds a four-layer governance model: technology, physical, logical and business. The agent diagnoses gaps between the layers, synchronises missing physical-layer assets and closes the cross-layer links.
A steward asks for a sales performance data product. The agent finds the right tables, validates the SQL, opens a Jira ticket to track delivery and registers the result as a data set in Collibra.
CAPABILITY LAYERS
{{product.name}} is organised into three layers. You work at the top, in Teams; the layers beneath handle the reasoning and the system connections.
Microsoft Teams is the single interface. Users ask in plain language, in the place they already work.
An AI agent runs a Thought, Action, Observation loop. It picks a tool, reads the result and repeats until it can answer or complete the task.
A tool server exposes more than 40 tools across Azure SQL, Collibra, Confluence and Jira. Each call goes to the live source system.
Every answer comes from the source system at the moment of the question. SQL access is limited to a single read-only query per call.
KEY CAPABILITIES
{{product.name}} is built and operated by {{product.owner}}. It works across the systems your data teams already use, so stewardship does not mean another place to look.
Every question is answered from Azure SQL, Collibra, Confluence or Jira at the time it is asked. There is no cache to go stale.
Runs on Azure, with a Model Context Protocol tool server, the Microsoft Bot Framework for Teams and a large language model reached through a managed gateway.
{{product.name}} reads and writes where your teams already work: tables in Azure SQL, assets in Collibra, pages in Confluence and tickets in Jira. You keep what works.
The {{product.name}} technology foundation
Ask for a walkthrough of {{product.name}} on your own data questions.
Book a session