XL
Spreadsheets → real schema
Years of Excel into a properly modelled, indexed Postgres database.
- Schema design
- Indexing
- Audit log
- Backups + PITR
Service · 03 / 06 · Foundations
Move the data. Keep the history.
From Excel, old CRMs, legacy databases and paper into clean, modern systems — without losing history.
- Typical project
- 3–8 wks
- History preserved
- Years
- Validation
- 100% reconciled
- Cut-over
- Reversible
Source → target
Five kinds of mess. One clean destination.
Most projects pull from 3–5 sources. We reconcile them into a single canonical model.
Sources
- SpreadsheetsExcel · Google Sheets · CSV
- Old CRMsSalesforce · HubSpot · Zoho · Custom
- Legacy databasesSQL Server · Oracle · Access · Filemaker
- Paper / manualOCR · form digitisation · review queue
- Disconnected SaaSStitch the 7 tools your team uses
Target
One source of truth
Postgres + clean schema, indexed for the reports you actually run.
- customersresolved
- orderslinked
- activity_feedpreserved
- attachmentsmigrated
Every row carries a reference back to its source so lineage is traceable forever.
The process
Five phases. One reversible cut-over.
- 01
Assessment
We open every source, count the rows and write the migration plan.
- 02
Mapping
Field-by-field mapping to a clean target schema.
- 03
Dry run
We migrate to a parallel environment. You verify before anything goes live.
- 04
Cut-over
Your team keeps working; we move history in the background.
- 05
Reconciliation
Every record is accounted for and the proof is in writing.
Capabilities
Everything you should expect, and most teams don’t get.
ID
Identity resolution
Dedupe at the identity level, not by name.
- Fuzzy match
- Manual review queue
- Merge history
- Survivorship rules
HX
History preservation
Activity history, notes and attachments follow the resolved record.
- Full activity feed
- Attachments
- Note threads
- Source references
VA
Validation & reconciliation
Row counts, checksums and sampled audits prove every record arrived.
- Row counts
- Checksums
- Sampled audit
- Discrepancy log
RB
Reversible cut-over
Dual-write windows, dry runs and a tested rollback plan.
- Dual-write window
- Tested rollback
- Zero-downtime option
- Phased cut-over
DC
Documentation
The new schema, mapping rules and migration log — handed over in writing.
- Schema docs
- Mapping rules
- Runbook
- Lineage map
FAQ
Questions, answered.
If something isn’t covered here, write to Info@corefluxsolutions.com — you’ll get a real answer within a working day.
How long does a typical migration take?
Tightly-scoped migrations land in 3–5 weeks. Multi-source consolidations with identity resolution are usually 5–8 weeks. We give you a fixed scope and date after assessment.
What if our source data is a mess?
That’s the normal case, not the exception. Assessment covers exactly that — duplicates, contradictions, missing fields. You see the problems before we touch anything.
Do you build the new system too, or just migrate?
Either. We migrate into your existing system if you have one, or we build the new home alongside — see ERP Solutions for the bundle.
Can we keep operating during the migration?
Yes. Dual-write windows or scheduled overnight cut-overs depending on risk tolerance. We’ve never had to halt a business to do this.
Who validates the result?
We do (with checksums and sampled audits) and you do (against records your team trusts). We don’t mark a migration complete until both sides sign off in writing.
What if it fails after go-live?
Reversible cut-over by default — we keep the old system warm for an agreed window. You can roll back without data loss.
Let's build
Have a system in mind? Let's sketch it together.
Tell us a little about the problem. We'll reply within one working day.
Write directly to Info@corefluxsolutions.com