Your files are not our product.
We keep the service simple: receive the agreed files, process them for the agreed purpose, review the result, return the deliverables and handle working copies according to the project's retention rules.
Only the data needed for the job.
We encourage clients to send the smallest dataset that allows the work to be completed. A product cleanup normally does not require unrelated customer records.
Access should have a reason.
Project files should only be available to the people performing or reviewing the specific job.
Unknown stays unknown.
Missing prices, SKUs, specifications or conflicting business facts are flagged for client confirmation rather than fabricated.
Automation is not the final authority.
AI or other automation may accelerate repetitive work internally. Final deliverables are reviewed against the agreed rules and QA checks.
Tool choice matters.
Third-party or AI tools should only be used for client data when appropriate for the task and consistent with the client's instructions and applicable project terms.
Working copies should have a limit.
We use a defined project retention period rather than treating client files as permanent archives. Longer retention can be agreed where necessary.
Material changes stay explainable.
Where appropriate, we provide a change/QA trail showing what was corrected, consolidated, flagged or left for review.
Problems get handled quickly.
Suspected unauthorized access or accidental disclosure should be recorded, contained and escalated using the project's incident procedure, including client notification where required.