Data Integrity & Architecture
Trusting the data you run the business on.
Workflow
Step 1Data Audit
ObjectiveAudit current data landscape and quality
InputData systems in use, current data
OutputData audit report with quality assessment
Key activities
Inventory data systems, assess data completeness, assess data accuracy, identify quality issues, identify gaps
Decision points
Data quality issues identified? Extent of problems? Priority fixes?
Tools: Data auditing tools, data quality toolsRoles: Data ops (audit), Analytics (assessment), IT (systems)
Success: Comprehensive data audit completed
Quality gate: Data audit completed and issues documented
Why it matters
- Every downstream decision — forecast, comp, pipeline — inherits data quality. Garbage data quietly invalidates the whole operating model.
Best practices
- Fix root causes, not symptoms — clean the source, not the spreadsheet.
- Put validation rules at ingestion.
- Assign data ownership per table.
- Audit on a cadence, and act on findings.
Common mistakes
- Endless "data cleaning" sprints that never fix the source.
- No owner — everyone's data, no one's data.
- Letting hygiene decay until forecast season.
Key questions
- Where does the dirt come from — tool, human, or process?
- Who owns this table's integrity?
Agents that drive this
Hygiene
Tools we use · alternates in [ ]
Supabase[ Postgres, Firebase ]n8n[ Zapier, Make, Airflow ]Metabase[ Looker Studio, Power BI, Grafana ]