| Cleaning | Sample files, a description of what “correct” looks like, known problem cases | Validation rules in writing, cleaned output, a rejection report per run | What to do with rejected records |
| Processing | Current manual steps, input locations, required outputs, timing | Automated job, run log, alerting on failure during working hours | Business rules and sign-off of first outputs |
| Hosting | Workload description, data sensitivity, budget range, any provider preferences | Provider options compared, configured environment, environment record | Choice of provider and data location |
| Backup | Recovery expectations: how much data you could lose, how fast you need it back | Retention pattern, configured copies, restore test notes | Acceptable loss and recovery time |
| Migration | Source and target systems, cut-over constraints | Migration plan, reconciliation counts, checksums, rollback route | Go / no-go at cut-over |