Skip to content
Vibedata

Work across more than one platform

Reconcile equivalent data across two live platforms before cutover

  • When two platforms are running the same logic in parallel and nobody has proven they agree.
  • When a cutover date is set but no one has reconciled the two estates row by row yet.

Reconcile equivalent tables and measures across two live platforms running in parallel, and classify every difference before cutover. Covers the reconciliation itself, not deciding which platform wins a disagreement.

Area
Data quality
Runs on
  • Microsoft Fabric Lakehouse
  • Microsoft Fabric Warehouse
  • MotherDuck
  • DuckDB
Built with
  • dbt
  • pytest
Readiness
SupportedEverything this Recipe composes runs today, without a case that proves this exact shape.

Sample This Recipe has not been materialized in the Cookbook repository yet. Its trigger, description, prompt, agent guidance and acceptance conditions, and the explanation below, are prototype drafts. Its name, job, area and readiness come from the reconciled Cookbook seed snapshot. Readiness is a separate question from this one: it says whether the capability exists, not whether the writing has been reviewed.

Use this Recipe

Use the VibeData Recipe `cross-platform-live-reconciliation` at https://getvibedata.ai/cookbook/cross-platform-live-reconciliation Read the Recipe and execute it in the context of the current Intent.

Recipe id cross-platform-live-reconciliation · Not yet materialized in the Cookbook repository, so the pointer addresses this page.

Verified by

What has to be observably true before this Recipe is finished.

  • row counts and key sets match between the two platforms for the agreed comparison window
  • every numeric difference above the agreed tolerance is classified with a cause
  • the reconciliation is repeatable and produces the same result on a second run
  • cutover is not recommended while an unclassified difference remains open
Recipe promptThe task specification the agent reads. Reference only — it is not what you copy.
Deliver: Reconcile equivalent data across two live platforms before cutover.

Reconcile equivalent tables and measures across two live platforms running in parallel, and classify every difference before cutover. Covers the reconciliation itself, not deciding which platform wins a disagreement.

Execute inside the current Intent. Its Domain, repository, platform, environment and attached sources are the context for this work — read them rather than asking for them.

The work is done when:
- row counts and key sets match between the two platforms for the agreed comparison window
- every numeric difference above the agreed tolerance is classified with a cause
- the reconciliation is repeatable and produces the same result on a second run
- cutover is not recommended while an unclassified difference remains open

Report the evidence for each condition above with the result. A condition you cannot meet is something to say, not something to work around.
Agent guidanceHow the agent approaches the work, and what it will not do.

Establish which platform is authoritative today and what in the project is platform-specific. Build against the target in an isolated copy and reconcile the two runs against each other before anything is cut over.

Composes

  • test authoring and baseline comparison
  • dbt in the project
  • pytest in the project
  • isolated-copy execution and gate verification

Asks first

Semantic decisions the Intent cannot supply. Never context Studio already holds.

  • which Fabric target this work lands on, when the Domain carries both a Lakehouse and a Warehouse

Guardrails

  • Build in an isolated copy. Production is read, never written.

What you need

  • Microsoft Fabric Lakehouse, Microsoft Fabric Warehouse, MotherDuck, or DuckDB.
  • The objects under test, and a baseline you agree is correct.
  • dbt and pytest in the project, or the intent to add it.

How it goes

  1. Name both platforms, and which one is authoritative today.
  2. Let it read what in the project is platform-specific and what is not.
  3. Review the plan before anything is rebuilt.
  4. Build and run against the target in an isolated copy.
  5. Reconcile the two runs against each other, at the grain the consumers read.
  6. Read the acceptance conditions before you cut over, not after.

What you end up with

The deliverable, in your own repository, as data quality work a reviewer who knows the project reads as native to it. Alongside it, the evidence for every one of the acceptance conditions above — which is the part that is still there in three weeks when somebody asks.

Next to this one