Regulatory compliance · Data quality

Turn regulatory requirements into enforced data quality checks.

Data quality for governance and platform teams under a regulator. The rows that fail a rule are kept as evidence.

Standardize checks across policies, regulations, and datasets

Give governance teams visibility into ownership, changes, and coverage

Meet compliance deadlines faster with business and engineering working together

“We just want to go and observe what’s happening.”

Sid Srivastava, 2K Games

When the regulator asks, what do you show them?

“We have to be able to prove that the copy is matching reality.”

“A spreadsheet which then relies on a human having to then interact with the actual physical data.”

“We have this data quality rules and then the question is, what do you do about it? I want to actually have a record.”

“You’re going to run into the same issue that we run into on everything, where no one knows who owns anything.”

1
2
3
4
5
6
7
8
9
10
11
dataset: .../agg_exposure_daily
checks:
  - freshness:
      name: Timeliness
      column: as_of_date
      threshold:
        unit: hour
        must_be_less_than: 4
      attributes:
        bcbs239:
          - P5

1
2
3
4
5
6
7
8
9
10
11
dataset: .../agg_exposure_daily
checks:
  - freshness:
      name: Timeliness
      column: as_of_date
      threshold:
        unit: hour
        must_be_less_than: 4
      attributes:
        bcbs239:
          - P5

An auditor does not want a green number. They want what failed, who was told, and that it was fixed.

What the team shows the auditor.

The rule, the sign-off, and the rows.

Every dataset in scope. One rule, written once.

✓

One list: the rule, its owner, every dataset it covers and each result.

✓

Write it once; every matching dataset inherits it. Public beta.

✓

Tag each rule with its regulation. Filter every result by that tag.

The steward drafts the rule. The owner publishes it.

✓

Propose it in the no-code editor or in plain English. No raw data is read.

✓

Every request keeps its author, status and date. The owner publishes it.

When a rule fails on rows, the rows are kept. In your warehouse.

✓

The rows that failed are stored with their check, scan and dataset tags.

✓

They live in your warehouse, not in Soda. The owner is notified by tag.

Their data stewards review the failed rows every week and get them remediated.

A data analyst at a utilities group, July 2026. Part of their data goes to councils under contract; the report of failures is the weekly worklist.

We’re creating a Power BI report on top of it to expose all those failed rows to the business so that they can work on remediation.

Data AnalystUtilities group

We went from detection to prevention. Most data quality tooling will tell you something broke after it broke.

Sid SrivastavaDirector of Data Governance, Quality and MLOps at 2K Games

Now it automatically runs all the checks we were doing manually and even stops report refreshes if something’s off, so the business never gets impacted.