Use case · Move off per-row pricing
Replace hosted ELT with self-hosted pipelines
Move pipelines off a hosted ELT service onto an orchestrator you run yourself — with a visual canvas, SQL and Python steps, resumable runs and per-environment licensing that does not grow with your row count.
- Pricing basis
- Per environment, never per row
- Connectors (open build)
- 100+ types
- Migration
- Incremental, side by side
- CDC
- Flow Enterprise (Debezium)
Who it is for
For data engineering leads whose hosted-ELT bill is now a line item the finance team asks about, and who want the migration to be incremental rather than a cut-over weekend.
What this looks like from the inside
Consumption pricing turned a routine backfill into a budget conversation, so the team quietly collects less data than it should. Meanwhile every sync copies production onto infrastructure nobody in the building controls.
Replacing it means replacing connectors, scheduling, retries and monitoring — not just the bill.
Drawn, not described
The one thing the prose cannot do: show the moving parts and where the boundary is.
hosted ELT
billed per row · vendor infrastructure
Revoluzen Flow
per environment · your infrastructure
- orders_dailyhosted sync retired
- stripe_refundshosted sync retired
- crm_contactsrunning both · diffing
- warehouse_inventorynot started
Overlap costs nothing on this side. Backfills on the right are not metered.
Step by step
In the order it actually happens.
- 01
Run Revoluzen Flow beside the existing service
Install the open-source build on a server next to the hosted tool. Nothing is switched off; the overlap period costs nothing on the Revoluzen side.
- 02
Rebuild one pipeline on the canvas
Pick a low-risk sync. Recreate it as a DAG — extract, conform, load — against the same destination table, using SQL or Python steps where the hosted tool had a black box.
- 03
Compare outputs, then switch the destination
Run both for a cycle. Diff the results. When they agree, point consumers at Revoluzen's output and retire the hosted sync for that pipeline.
- 04
Repeat, and let backfills stop costing anything
Per-step resume and idempotent backfills mean a failed run continues from the failed step and a re-processed window does not duplicate rows. Backfill freely; the license does not meter it.
- 05
Bring in CDC and the enterprise sources when you need them
Change data capture via Debezium and the systems of record — SAP, Oracle, Salesforce, Workday — are Flow Enterprise. Add them when a pipeline needs them, not before.
What actually changes
| Dimension | Today | With Revoluzen |
|---|---|---|
| What drives the bill | Rows processed. A backfill or a growth month is a cost event. | The number of environments. Volume does not change the number. |
| What a pipeline looks like | A vendor's configuration screen; transformations somewhere else. | A DAG you can read — extract, conform, assert, load — with SQL and Python in the same graph. |
| What happens on failure | Re-run the sync from the start and hope the destination tolerates duplicates. | Resume from the failed step; backfill idempotently; see the step that failed and why. |
Which product does what
Each product is adoptable on its own; together they share one catalog, one permission model and one run history.
Revoluzen Flow
Visual DAG canvas with 40+ step types plus SQL and Python; 100+ connector types; per-step resume and idempotent backfills so a re-run does not duplicate rows.
Product pageRevoluzen Flow Enterprise
Adds change data capture via Debezium and the enterprise systems of record — SAP, Oracle, Salesforce, Workday and others — plus workspaces and promotion between environments.
Product pageWhere it runs
Start on one laptop with the open-source build and a single real source. Move to an internal server with a Postgres metastore when a team depends on it. Nothing structural changes between the two.
What you end up with
- 01Pricing by environment: backfills and growth do not change the bill
- 02Pipelines you can read — a canvas, not a vendor's configuration screen
- 03Steward reliability findings — 37 kinds — before a bad run reaches the warehouse
- 04A migration you can run incrementally, one pipeline at a time, side by side
The honest limits
You would find these out eventually. Finding them out now is cheaper for both of us.
Connector maturity is published per connector; none is yet marked production-ready. Check the ones you depend on before planning a cut-over.
Change data capture and the enterprise systems of record are commercial (Flow Enterprise), not in the open-source build.
A hosted service's long tail of niche SaaS connectors may not be matched one for one; the catalog is the place to check.
Asked about this scenario
Is Revoluzen Flow a self-hosted alternative to hosted ELT services?
Yes. Revoluzen Flow is an Apache-2.0 orchestrator that runs on your own infrastructure with a visual canvas, SQL and Python steps, scheduling, retries and resumable backfills, and it is licensed per environment rather than by rows processed.
How do I migrate pipelines from a hosted ELT tool to Revoluzen?
Incrementally: run Revoluzen Flow next to the existing service, rebuild one pipeline at a time against the same destination, compare outputs, then retire the hosted sync. The open-source build is free, so the overlap period costs nothing on the Revoluzen side.
Does Revoluzen Flow support change data capture?
Change data capture via Debezium is part of Revoluzen Flow Enterprise, the commercial edition. The open-source build covers batch and incremental extraction with watermarks and lookback windows.
Adjacent scenarios
Keep data in-country, in-network
Data residency and sovereignty
Run pipelines, a lakehouse and analytics entirely inside your own network — with no hosted control plane, no telemetry and no outbound dependency — so residency obligations are met by architecture rather than by contract.
Medallion, modelled, enforced
A governed lakehouse you operate
Build a bronze–silver–gold lakehouse on open table formats and your own object storage, with contracts enforced between layers, dimensional models as first-class objects, and a SQL endpoint that applies row-level security.
Run it yourself, then decide.
Revoluzen Flow is Apache-2.0 and installs in one command. If it fits, we can talk about the parts that are not open source.