# CRM Contacts Change Summary

Public synthetic sample for Rowva artifact inspection.

## Source

- File: `crm_contacts_messy.csv`
- Source rows: 8
- Operational export rows: 6
- Cell-level fixes logged: 24
- Rows excluded from operational export: 2
- Conflicts preserved for review: 4

## Automatic fixes

- Normalized contact IDs to `c_###` format.
- Canonicalized company names where duplicate evidence was strong.
- Standardized regions such as `NAM`, `NAMER`, and `North America`.
- Converted mixed date formats to ISO dates when unambiguous.
- Removed summary row from clean record export.

## Review items kept

- Invalid email value for row 3.
- Missing email for row 4.
- Rows 5 and 6 have same email and similar company/name values.
- Phone numbers and names remain sensitive and should be excluded from AI packages unless policy allows.

## Readiness

- Operational Clean Export: partial but usable with review.
- Compliance-Safe Export: ready after phone numbers and notes are excluded.
- Reporting Package: partial because duplicate contacts and invalid emails remain.
- AI Package: blocked because direct contact details remain.
- Audit Trail Bundle: available.