A workshop of short, hands-on modules for data engineers and analytics engineers who need to trust — and prove — answer accuracy. The thesis throughout: accuracy is checked, not asserted. Golden datasets pinned to numbers the org already trusts, eval cases that exercise edge conditions, drift detection on a schedule, reconciliation workflows for when systems disagree, and a validation agent that runs the moment data lands.
Total time: about 2 hours.
Builds on Build Your Ontology, End to End and Ontology Operations — this workshop assumes governed definitions exist and doesn’t re-teach the build; it teaches the proof.
Who this is for
- Data/analytics engineers accountable for “is this number right?”
- Ontology owners who need tests, not vibes, behind their definitions
- Teams burned by a silent upstream change that shipped wrong numbers for a week
- Anyone preparing for an audit, a board cycle, or a regulator
What you’ll be able to do
| After module | You can… |
|---|---|
| 0 | Map every way an answer goes wrong to the defense that catches it |
| 1 | Build golden queries pinned to numbers the org already trusts |
| 2 | Generate and curate eval cases that exercise definitions and edge cases |
| 3 | Run drift detection on a schedule, distinguishing data drift from definition drift |
| 4 | Reconcile two disagreeing systems to the exact fork, and record the resolution |
| 5 | Deploy a data-quality agent triggered by ETL completion |
| 6 | Write the quality runbook: owners, cadences, and what a red alert triggers |
Workshop modules
| Module | Time |
|---|---|
| 0 · The Trust Stack | 10 min |
| 1 · Golden Datasets | 20 min |
| 2 · Validation Sets & Eval Cases | 20 min |
| 3 · Drift Detection on a Schedule | 20 min |
| 4 · Reconciliation Workflows | 20 min |
| 5 · The Data-Quality Feed Agent | 15 min |
| 6 · The Quality Runbook | 15 min |
How to use this workshop
- Prerequisites: a connected warehouse and an existing ontology with governed metrics (the build workshops above if not).
- Work in order — golden datasets (Module 1) are the foundation every later module schedules, extends, or operationalizes.
- Swap
[bracketed]placeholders for your metrics, tables, and trusted sources.
For facilitators — running this as a guided session
For facilitators — running this as a guided session
T-minus-1-day pre-flight: run the workshop’s key prompts end-to-end in the session workspace — confirm logins, connectors, and the features this workshop touches are enabled for every attendee; have a fallback demo workspace ready in case a customer connector fails live. Pacing: when a long prompt is running (2–4 min), fire it first, then discuss the concept while Ana works — never watch a spinner in silence. If behind schedule, cut sections marked optional, never the checkpoints. Group sessions: attendees drive, you narrate; collect every miss or wrong answer in a shared doc — that list is the ontology backlog. With 10+ attendees on one warehouse, stagger the heavy prompts.