Cataloging 1,600 Undocumented Tables in 3 Days for a Post-Acquisition Retail Integration
Metadata discovery completed
Source estate
Documented
Delivered within hours of the request
THE CHALLENGE
An Australian retailer had acquired a competitor and needed to integrate the acquired company's data estate, which ran across eight MySQL databases. The estate was undocumented, and no subject matter experts were available to explain it. The target was Google BigQuery, and the integration could not proceed until the team understood what the acquired estate actually contained.
THE SOLUTION
The engagement applied Metadata Intelligence and Reverse Engineer accelerators directly against the undocumented sources. Because no SMEs were available, discovery could not rely on interviews. It had to read the systems themselves: extracting metadata, profiling the data, and recovering structure and meaning from the databases as they existed.
WHAT WAS DELIVERED
- 1,600 tables and more than 24,000 columns documented.
- More than 50 candidate KPIs delivered within hours of the request.
- A BigQuery target architecture unblocked so the integration could proceed on schedule.
- A permanent metadata asset retained for ongoing governance.
WHY IT WORKED
The constraint was time and the absence of tribal knowledge, and automated metadata discovery addressed both. Cataloging 1,600 tables and 24,000 columns by hand would have taken weeks the integration did not have. Reading the estate directly compressed that to three business days and produced a documented foundation the team could design against immediately.
THE LASTING VALUE
Beyond unblocking the immediate integration, the catalog became a durable governance asset. The knowledge that did not exist in any person now existed in a maintained form, which is exactly what an organization needs after an acquisition when the people who understood the acquired systems are not part of the team.
ACCELERATORS USED
• Metadata Intelligence: Source-connected metadata extraction, data profiling, and documentation across undocumented MySQL sources.
• Reverse Engineer: Structure and meaning recovery from the acquired databases when SMEs were unavailable.
KEY TAKEAWAY
Need to catalog an undocumented source estate before post-acquisition BigQuery integration?
3X Data Engineering read the acquired estate directly across eight MySQL databases, extracted metadata, profiled the data, and recovered structure and meaning without relying on SME interviews.
The result was a documented metadata foundation covering 1,600 tables and 24,000+ columns in 3 business days, with 50+ candidate KPIs delivered, BigQuery target architecture unblocked, and a permanent governance asset retained.