Migrate ERP data you can see before you move it
Design the migration on a canvas, preview exactly what lands in the destination, then run it on a queue that logs every record it touches.
Your own PostgreSQL schema. Ready in about a minute.
Source tables
| order_no | customer_id | status |
|---|---|---|
| SO-10041 | NORDWIND | Released |
| SO-10042 | CONTOSO | Planned |
| SO-10043 | FABRIKAM | Released |
Built for the parts that usually break
Schema drift, silent truncation, a run that dies halfway. The tool is shaped around the failures, not the happy path.
Visual flow designer
Map source tables to destination entities on a canvas. Middleware nodes hold the filters, joins and derived columns in between.
Prebuilt flows
Every workspace is seeded with ready-made migration templates. Clone one, point it at your data and adapt from there.
Queue-backed runs
Migrations execute on a worker queue with live progress, automatic retries and a dead-letter lane you can requeue from.
Isolated per tenant
Each workspace gets its own PostgreSQL schema, so your tables never share space with another organisation.
AI-assisted mapping
Field mapping suggestions and prerequisite checks, offered before a run so you catch the gaps rather than discover them.
Work tracking
Plan the cutover in the same place you run it: assignments, priorities, workflow stages and approvals.
What you actually work in
Not a config file and a cron job. A workspace built around the day-to-day of moving ERP data and living with the result.
Browse the IFS entities, see row counts and how stale each one is, then add the ones you need to a flow.
CustomerOrder
12,480 rows| order_no | customer_id | state | wanted_delivery |
|---|---|---|---|
| SO-10041 | NORDWIND | Released | 2026-09-12 |
| SO-10042 | CONTOSO | Planned | 2026-09-14 |
| SO-10043 | FABRIKAM | Released | 2026-09-15 |
| SO-10044 | TAILSPIN | Invoiced | 2026-09-18 |
Tenant isolation is the default, not a tier
This is production ERP data. The separation, the encryption and the audit trail are there on the free plan too.
Schema-per-tenant
Your tables live in their own PostgreSQL schema, resolved from your subdomain on every request.
Secrets encrypted at rest
IFS client secrets and session cookies are encrypted in the database, not stored in plain text.
SSO with Google and Microsoft
Sign in through your existing identity provider. Tokens stay in httpOnly cookies, never in browser storage.
Roles and approvals
Scoped permissions per user, with an approval step on the actions that warrant one.
Audit trail
Activity logs record who ran what and which records moved, kept per tenant.
Rate limited by default
Auth and expensive endpoints sit behind separate limits, so one caller cannot starve the rest.
Three steps to a first migration
- 1
Connect IFS
Point the tool at your IFS OData endpoint and pick the entities you actually need.
- 2
Design the flow
Drag source tables onto the canvas, add the transforms, and preview the result before anything moves.
- 3
Run and monitor
Queue the migration, watch progress live, and review every touched record in the activity log.
Two ways to start
Begin free and self-serve. Talk to us when the volumes outgrow it.
Free
Start on your own. Seat and usage limits apply.
- Self-serve signup, no sales call
- Your own isolated PostgreSQL schema
- Prebuilt migration flows included
Enterprise
Higher limits and a migration plan built with you.
- Volume and seat limits raised to fit
- Onboarding with the Change8 team
- Priority support
Questions we get asked
- Which IFS versions does it connect to?
- Anything exposing IFS OData. You point the workspace at your OData endpoint and pick the entities you need — source and destination are configured separately, so you can move between two IFS environments as well as into PostgreSQL.
- Where does my data actually live?
- In its own PostgreSQL schema inside the platform database, resolved from your subdomain. Nothing is shared with another workspace, and the migrated data stays queryable through virtual tables and custom views.
- What happens when a migration fails halfway?
- The run is queued work, not a browser session. Failed jobs retry automatically and then move to a dead-letter lane you can inspect and requeue. The activity log keeps a record of everything that was touched before the failure.
- Can I see what a migration will do before running it?
- Yes — that is the point of the canvas. Middleware nodes can be previewed against real data, so you see the output rows before anything is written to the destination.
- Do I need a card to start?
- No. The Free plan is self-serve with seat and usage limits, and there is no payment step. When you outgrow it, you talk to us.
- What if I need a workspace for more than one company?
- Create one per company. Each gets its own subdomain, schema and prebuilt flows, and the same email address can hold a seat in more than one.
Start with your own workspace
Pick a subdomain, verify your email, and the schema plus prebuilt flows are provisioned for you.
Create your workspace