Move the Data.
Keep the Logic.
Migrations stall on rewriting logic and proving the new numbers match. With Prometheux, business rules are platform-neutral: point them at the new platform, run old and new side by side, and switch once the results agree.
revenue_legacy✓revenue_new✓parity_diff✓Proven in production
Moving Data Is
the Easy Part.
Copying tables is rarely what takes months. Rewriting the logic tied to the old platform, then proving the new numbers match the old ones, is.
Change Platforms.
Keep Your Logic.
Because every rule refers to concepts in your ontology, not to a vendor's SQL dialect, the logic does not need to be rewritten when data moves. You point them at the new sources, and the same rules run against the new home.
Repoint, Don't Rewrite
Point a concept at its new location and every rule built on it keeps working.
Declarative Data Moves
Read from the old system and write to the new one in a rule, with append or overwrite.
Test Before Cutover
Run the same rules against both systems and compare the outputs value by value before, during and after the move.
Across Any Connector
Move between relational databases, warehouses, lakehouse formats like Iceberg, and files.
Shrink migration timelines and risk. Stop paying to rebuild logic you already have. Make platform decisions on merit, not switching cost.
A Migration in
Four Steps
Model Once
Capture entities and business rules once, independent of where data sits.
Connect Old and New
Connect both the source and the target platform at the same time.
Move With a Rule
Read from one, write to the other, transforming on the way if needed.
Compare, Then Switch
Run the same rules on both sides, check the results match, then switch over.
Moves That
Used to Stall
Move to Databricks Gradually
Analyze, test and transform data before, during or after migration, with no re-architecture or downtime.
Prometheux on Databricks →Off the Mainframe
Read COBOL and legacy relational data and write it to modern formats with the same rules.
Follow Your Credits
Move data between clouds when pricing, credits or regional needs change, without rebuilding pipelines.
No Lock-In case study →Merge Data Estates
Map both companies' systems to one shared model first, then consolidate at your own pace.
“Generative AI without ontologies is an engine without a road. Prometheux builds the road.”
Marco RomeiDeputy CIO · Banca Sella
Questions,
Answered
Do I have to migrate to use Prometheux?
No. Prometheux works across your existing databases, warehouses and platforms without mandatory migrations. When you do want to move data, it makes the move simpler.
How is this different from automated code translation?
Translation tools, including AI agents, convert old SQL or pipeline code into the new platform's dialect, which then has to be reviewed, tested and maintained. Prometheux skips the translation: logic lives in platform-neutral rules, so you repoint the data and keep the rules you already trust.
What about the SQL we already have?
Rules can embed SQL bodies directly, so existing queries can be brought in as they are and refactored into shared concepts over time.
Which systems can I migrate between?
Read from any documented connector, including relational databases, Snowflake, Databricks, BigQuery, Redshift, Iceberg, Parquet and CSV, and write to any that supports writing, such as PostgreSQL, SQL Server, Snowflake, Databricks, Iceberg, Parquet and CSV.
Can I validate before cutting over?
Yes. Connect both the old and the new system, run the same rules against each, and join the two outputs to flag any value that differs. Rules are deterministic, so a difference points to the data, not the logic. Put the check on a schedule to keep watching until cutover.
Related Capabilities
Plan a Migration
Without the Rewrite
Tell us what you're moving and where. We'll show you the rules that move it.
