Public-source data methodology.
DModify separates authoritative source retrieval from normalization, validation, provenance and controlled delivery.
1. Retrieve
Records are collected from identified public source systems using repeatable production workflows.
2. Normalize
Source-specific structures are mapped into stable, machine-readable dataset contracts.
3. Validate
Required quality rules separate accepted delivery records from QA exceptions.
4. Deduplicate
Production state helps prevent unintended redelivery and supports version-aware processing.
5. Reconcile
Dataset-specific relationships and monetary histories are reconciled where required.
6. Deliver
Accepted records can be packaged with schema versions, counts, provenance and integrity metadata.
Methodology is separate from record-level quality.
Quality describes the validation state of records and deliveries. Methodology describes how source data moves through acquisition, normalization, provenance preservation, validation and controlled delivery.
Source traceability
Where supported by the source, identifiers and provenance metadata are retained so delivered records can be traced back to authoritative public systems.
Validated production runs
Delivery is based on validated production outputs. Failed or invalid production runs are not silently substituted for accepted output.
Read-only delivery boundary
Delivery services consume validated production outputs without modifying the underlying production source files.