We assess read-only database access, exports, reports, backups, files, event logs, existing integrations, and controlled extraction methods. The chosen path must be legal, secure, repeatable, and reconcilable. Unknown fields, relationships, encodings, attachments, and business rules become explicit discovery items.
Build a repeatable extraction path before estimating the move
An API is one way to access records, not the only one. Assess authorised exports, reporting tools, read-only database access, backups, files, and existing integration feeds. Prefer a method that can be repeated safely and that preserves identifiers and relationships. Do not experiment against a live production database without appropriate controls.
Use a small sample to check character encoding, timestamps, currencies, attachments, and references between records. Then determine how changes since the first extraction will be identified for cutover. If extraction is a one-off manual process, make its cost, limitations, and final-update plan explicit rather than assuming it can be repeated on launch day.
What to check before deciding
- Document permitted read access and the extraction owner.
- Validate identifiers, relationships, encoding, and media.
- Test repeatability and reconciliation on a representative sample.
- Define how new and changed records will be captured before launch.
