What we solve ·What is unaccounted for, and what am I missing?
Forge · Build
Will your data hold what you are about to build on top of it?
Every area has its own version of the same figure and none of them match.
Whether your data can carry what you are about to buy, measured against a rubric and with the cost of closing each gap.
Data foundations auditSend this page to whoever decides
What you receive
- The domain quality report with a score per dimension.
- The inventory of silos and conflicting definitions.
- The data ownership map, showing what has an owner and what does not.
- The prioritized gap report and the phased remediation plan with an order of magnitude.
The proof that applies here
- The separation of two banks' data in one of the largest bank divestitures in Latin America: 100% of systems cut over on the legal day, zero failures and zero regulatory findings.
- Distributed infrastructure of 60,000 servers.
- Eight years of reporting under supervision without a single finding, which is a statement about the data before it is one about the process.
How we solve it
The method, not the promise.
- One domain gets scoped. Auditing everything at once is the most common way of never finishing.
- It gets measured against the six dimension rubric, with data and not with interviews.
- The silos get hunted: which figure exists twice, with what definition and who defends each version.
- Ownership gets mapped, which is the root cause of everything else.
- It gets prioritized by what each gap prevents you from doing, with the specific initiative named beside it.
Use this today, without hiring anyone
The six dimensions used to measure the quality of a domain. Pick the domain that worries you most (customers, products, contracts) and score it from one to five on each. In one afternoon you have your first picture.
- Completeness. What percentage of records have the fields that matter filled in. Count, do not estimate.
- Accuracy. Take fifty random records and verify them against reality. It is tedious and it is the only one that cannot be simulated.
- Timeliness. How long it takes for an event in the world to appear in the system. If it is longer than the decision lasts, the data arrives late even when it is correct.
- Uniqueness. How many times the same customer or the same product appears. Look for duplicates on two different fields and compare.
- Consistency across sources. The same figure in two systems: does it match? If not, which one governs and is that written down?
- Lineage. Can you say where one specific number in your report came from? If not, any argument about it is unresolvable.
Behind all six dimensions there is a single root cause. Who owns that domain? Not which area uses it: who decides what is correct. If nobody holds that authority, no vendor can solve it for you.
This sounds like you if
- A project stalled the moment it reached the data.
- The same indicator arrives from two areas with different definitions.
- A vendor made data quality a precondition.
Before you hire
The audit stands on its own and ends in week three. Remediation gets quoted by phases and requires assembling a technical team: it is not sold as installed capacity.
What is scarce is not auditing data against a rubric: it is that whoever audits knows what breaks downstream when a dimension fails. That is learned by separating two banks' data against a contractual date.
In your most critical domain, who decides today what data is correct?
If the problem is a different one
The committee gets forty pages every month and leaves without deciding anything.
Decision grade reporting
By the time the financial number turns red, the problem is four months old.
Leading indicator scorecard
I know there is AI in use and I do not know which tools or with what information.
Shadow AI inventory