Skip to content

[Data Quality] Add Great Expectations checks for silver and gold layers #2

@Rashmi-278

Description

@Rashmi-278

Context Update (2026-02-26)

PR #1 moved silver/gold transforms to dbt, which now provides 120+ schema tests
(not_null, unique, accepted_values) for silver and gold models. This significantly
reduces the urgency of adding GE checks for these layers.

GE checks for silver/gold would provide defense-in-depth validation beyond what
dbt tests cover (e.g., distribution checks, statistical expectations, cross-table
consistency).

Original Problem

The bronze layer has full GE validation coverage (5 suites), but silver and gold
layers lack any Great Expectations checks.

Remaining Scope (reduced)

  • Add GE suites for gold models only (governance_activity, treasury_summary,
    delegate_scorecard) where statistical/distribution checks add value beyond dbt tests
  • Focus on checks dbt tests can't easily express:
    • Row count ranges (not just > 0)
    • Column value distributions
    • Cross-table consistency (e.g., all addresses in gold exist in address_crosswalk)

Priority

P3 (Low) — dbt tests now provide primary coverage. GE adds defense-in-depth.

Files to Modify

  • infra/validate/checks.py — Add new GE check functions
  • infra/great_expectations/expectations/ — New suite JSON files
  • infra/definitions.py — Register new checks

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions