Set the reconciliation question
Configure an appropriate check for the source-to-target workflow being evaluated.
Data reconciliation for operational workflows
Prove each data load arrived as expected.
Recon gives data teams a repeatable way to compare expected source and target outcomes around a load or replication event. It supports Oracle, SQL Server, and Databricks in a controlled deployment.
Reconciliation path01 - 03
01
Define
Configure the check for the source-to-target workflow.
02
Check
Compare expected source and target outcomes around the workflow.
03
Review
Begin the approved investigation process when a difference is found.
Current adapters
- Oracle
- Microsoft SQL Server
- Databricks
The question after a completed job
A job running is not the same as data agreeing.
Pipelines can complete while data is incomplete, duplicated, or changed. Recon helps teams investigate whether a load or replication event produced the expected result.
Compare the expected outcome with the reported outcome, then give the difference a defined place to be investigated.
A reconciliation question
Expected source outcomeConfigured check
Reported target outcomeReviewable result
How Recon works
A bounded check around the workflow you already operate.
Recon fits around an existing load or replication process. It does not replace the systems that move or transform data.
Compare outcomes
Run the check around the existing workflow to compare expected source and target outcomes.
Investigate a difference
Review the result and follow the team's approved process when something does not match.
Configured checks
Compare what matters to the workflow.
The appropriate reconciliation check depends on the data workflow and the question the team needs to answer when something does not match.
Counts
Confirm expected record volumes for a complete load at a high level.
Key coverage
Identify gaps or duplication in configured records to investigate coverage and key integrity.
Selected values
Compare configured data values where a targeted data difference needs investigation.
Available checks are confirmed during evaluation.
Built for controlled operations
A clear role in a mature data workflow.
Recon is intended for controlled deployment in the environment chosen for the data workflow, with results review limited to the teams approved for the deployment.
Fits around existing operations
Use Recon around the load or replication processes a team already operates.
Focused review
Results give the responsible team a clear outcome and a defined starting point for investigation.
Evaluation-led fit
Technical evaluation confirms the appropriate checks and requirements for the intended workflow.
Current supported source and target adapters
- Oracle
- Microsoft SQL Server
- Databricks
Start with one representative workflow
Map the check, outcome, and evaluation goal in a technical walkthrough.
Bring one representative data workflow, the expected checks, and the team that will evaluate Recon.
- Which systems does Recon support today?
- Does Recon replace a data pipeline?
- What happens when Recon finds a mismatch?