DEA Level 3 case studies
DEA Level 3 · Put your knowledge to work in context.
The DEA Level 3 is an advanced, multiple-choice certification aimed at data professionals who already work across the data lifecycle and want to demonstrate senior-level breadth. It suits data engineers, analytics leads, data architects and governance specialists who are comfortable with both hands-on technical work and the organisational side of data. The DEA Level 3 case studies are 5 cases and about 3,197 words.
1The Reconciled Customer Count That Nobody Trusts
A retail bank's Chief Data Officer is told by the CFO that three reports circulated in the same week show 4.12M, 4.38M, and 4.61M active customers. Each report was built by a different team: Finance uses the general ledger's customer-of-record flag, Marketing uses the CRM's active-account definition, and Risk uses any customer with a transaction in 90 days.
628 words, about 3 min2The Lakehouse Migration With a Frozen Reporting Layer
A logistics firm is migrating from an on-premises warehouse to a cloud lakehouse. The legacy warehouse has 1,800 tables, of which 240 feed regulatory filings and 60 feed a customer-facing SLA dashboard with a 15-minute freshness requirement.
607 words, about 3 min3The Nightly Batch That Cannot Meet the Fraud Window
A payments processor detects fraud with a nightly batch that scores the prior day's transactions. Chargebacks are rising and the risk team says decisions must be made within 2 seconds of authorization.
655 words, about 3 min4The Churn Model That Passed Validation and Failed in Production
A subscription business deployed a churn model that achieved 0.86 AUC in offline validation. Six months later, the retention team reports it is 'useless' - it flags mostly customers who were never at risk, and misses the high-value accounts that actually leave.
661 words, about 3 min5The Consent Gap in a Cross-Border Analytics Program
A healthcare-adjacent company runs analytics on user behavior across the EU, UK, and US. A new program would combine EU app-usage data with US purchase history to build a propensity model.
646 words, about 3 min