Features

Everything you need.
Nothing you don't.

Define contracts in YAML. Sync to dbt. Validate in CI. Block bad data. Report to stakeholders.

Define

YAML contracts, ODCS v3.1 compatible

Readable. Versionable. Standard. Define schemas, quality rules, SLAs, and ownership in a single file — no Python required. DataVow implements a superset of the Open Data Contract Standard.

✓ Schema validation (types, required, unique, patterns)
✓ Custom SQL quality rules
✓ Freshness SLAs
✓ Domain ownership metadata
✓ PII field tagging

# contracts/sales/orders.yaml
apiVersion: datavow/v1
kind: DataContract
metadata:
  name: orders
  domain: sales
  owner: data-team@company.com

schema:
  fields:
    - name: order_id
      type: integer
      required: true
      unique: true
    - name: customer_email
      type: string
      pii: true

quality:
  rules:
    - name: no_negative_totals
      severity: CRITICAL
      

Sync New

Contracts become dbt tests

datavow dbt sync reads your YAML contracts and generates dbt generic and singular tests. One source of truth — write the contract once, the tests follow.

✓ 3 contracts → 28 tests generated automatically
✓ Schema rules → dbt generic tests (not_null, unique, accepted_values)
✓ Custom SQL rules → dbt singular tests
✓ --clean flag to rebuild from scratch

Validate & Block

Bad data doesn't reach production

DataVow runs as a dbt post-run hook. When tests fail, the pipeline is blocked with a clear message: which tests failed, what the Vow Score is, and what to fix.

11 failures? Score: 0/100. Pipeline blocked. No ambiguity — fix the violations or explicitly accept them.

CI/CD New

GitHub Action — automated enforcement

datavow-action@v1 runs on every push and PR. It validates all contracts, generates HTML reports, uploads them as artifacts, and fails the build if any contract is broken.

✓ orders_clean: Vow Kept (score=95)
✗ orders_dirty: Vow Shattered (score=15)
✓ Reports uploaded as CI artifacts
✓ Exit code 1 → merge blocked

Comparison

Built for the gaps others leave.

DataVow isn't another observability platform. It's the missing enforcement layer.

	dbt tests	Great Expectations	Soda	DataVow
Contract format	SQL only	Python	YAML	YAML (ODCS)
dbt integration	Native	Manual	Plugin	Auto-sync
Learning curve	Low	Steep	Moderate	Low
Audience	Engineers	Engineers	Eng + analysts	Whole team
CI/CD native	Via dbt	Manual	Manual	GitHub Action
Stakeholder reports	No	No	Limited	HTML reports
Data Mesh ready	No	No	Limited	By domain
Pipeline blocking	On failure	Manual	Manual	Score-based
ODCS standard	No	No	No	Yes

Everything you need.Nothing you don't.