Route data-quality failures to the right owner
Failures are grouped by source, impact, recurrence, and accountable steward with a clear next action.
The problem
Broken fields and stale pipelines create downstream confusion without a shared severity or owner model.
What you get
A working system with the steps, tools, checkpoints, and expected return made explicit.
- Setup
- 75 minutes
- Back each week
- 3 hours
- Difficulty
- agentic
Expected return
The working case
A planning estimate, not a guaranteed result. Measure the first four weeks against your own baseline.
144
hours returned per year
At 3 hours/week across 48 working weeks.
1
week to earn back setup
Compare the setup estimate with the weekly time returned.
Operating contract
Input
The source material, constraints, and examples a human would need to do this work well.
Checkpoint
A person reviews judgment calls, sensitive content, unfamiliar tools, and irreversible actions.
Success signal
Track time returned, corrections required, and exceptions. Keep it only if the measured result compounds.
Before you start
- ·A named data operations lead
- ·An approved source-of-truth and review template
The steps
- 01
Name data operations lead as the accountable owner and define the decision this workflow is allowed to support.
- 02
Collect validation results, lineage, incident history, and data ownership; preserve source links, timestamps, and access controls before any synthesis.
- 03
Produce a prioritized data-quality exception queue using the approved template. Do not repair production records automatically.
Copy this prompt
Create a prioritized data-quality exception queue from the supplied evidence. Separate facts, assumptions, and missing inputs. Cite every material claim. Do not repair production records automatically. Evidence: [approved inputs]
- 04
data operations lead reviews the draft, records the decision or next action, and corrects the source system before distribution.
What it runs on
Where this goes wrong
- Do not let the model act beyond do not repair production records automatically.
- Keep sensitive fields out of unapproved tools and retain a human-readable evidence trail.
Definition of done
Run it for four weeks. Then make it earn its place.
- □ Baseline the manual time before launch.
- □ Keep a human approval step for consequential output.
- □ Record corrections and exceptions, not just successes.
- □ Expand, revise, or retire it after the first review.
Build the system around it
Related workflows
If this one stops working, tell us. Three reports in a month and it leaves the library until a person has looked at it again.