22-25 April 2026 
Autonomous Fabric: Agentic Data Quality Using Fabric Data Agents
Proposed session for SQLBits 2026TL; DR
What if Fabric Data Agents could detect Data Quality issues autonomously and guide engineers directly to a fix?
This session demonstrates capturing DQ signals, triggering agent-led investigations, and integrating agent outputs into ETL pipelines.
Session Details
Struggling with Data Quality issues?
Manual checks, slow root-cause investigations, and hard-to-diagnose issues consume engineering time that should be spent delivering value.
To address these challenges, this session demonstrates how Fabric Data Agents provide an intelligent diagnostic layer for automated Data Quality operations.
You’ll see how Data Quality signals get captured, are routed to a Fabric Data Agent, and used to drive targeted investigations and potential fixes.
Tailored for Data Engineers and Data Architects, a live demo will showcase the full flow: we’ll introduce a real data issue into a Fabric Lakehouse table, run a lightweight Data Quality check, and watch the Data Agent in action as it:
• Reasons about data anomalies using AI
• Identifies the likely root cause
• Generates diagnostic SQL
• Provides clear, actionable recommendations
Attendees will leave with:
• A repeatable agent architecture for automated Data Quality diagnostics
• Practical guidance on why this pattern delivers real value
• Techniques for integrating agent-driven insights into existing engineering workflows
Manual checks, slow root-cause investigations, and hard-to-diagnose issues consume engineering time that should be spent delivering value.
To address these challenges, this session demonstrates how Fabric Data Agents provide an intelligent diagnostic layer for automated Data Quality operations.
You’ll see how Data Quality signals get captured, are routed to a Fabric Data Agent, and used to drive targeted investigations and potential fixes.
Tailored for Data Engineers and Data Architects, a live demo will showcase the full flow: we’ll introduce a real data issue into a Fabric Lakehouse table, run a lightweight Data Quality check, and watch the Data Agent in action as it:
• Reasons about data anomalies using AI
• Identifies the likely root cause
• Generates diagnostic SQL
• Provides clear, actionable recommendations
Attendees will leave with:
• A repeatable agent architecture for automated Data Quality diagnostics
• Practical guidance on why this pattern delivers real value
• Techniques for integrating agent-driven insights into existing engineering workflows
3 things you'll get out of this session
1) A repeatable architecture for automated Data Quality diagnostics using Fabric Data Agents, from DQ signal capture to ETL integration.
2) Practical techniques demonstrated in a live scenario where an agent identifies problem root causes, generates diagnostic SQL, and recommends fixes.
3) Clear guidance on why this pattern delivers value and how to adopt agent-driven diagnostics in engineering environments.