PART OF THE EBS → FUSION CASE STUDY

    Row- & Column-Level Reconciliation for the Oracle Fusion Migration

    Reconciliation and validation were not an afterthought — they were the deliverable. Every object produced a source → staged → Fusion count and row- and column-level validation of every data element, backed by load logs and a shared error tracker.

    Source→staged→Fusion
    Per object
    Row + column
    Every element
    Load logs
    Success/fail counts
    Error tracker
    Every exception

    Two kinds of validation, kept separate

    🎯

    Load-readiness validation (Syntra)

    Before data reaches Fusion, Syntra checks mandatory fields, datatypes, lookups and referential integrity against the Fusion targets. Records that would fail import are caught in staging and reported back — not discovered mid-load.

    🏛️

    Business validation (PwC)

    Business correctness and cleansing of source data sat with PwC. Keeping the two apart meant every reject had a clear owner and a clear fix.

    The reconciliation pack, per object

    For every object, Syntra produced a technical reconciliation showing three counts and the deltas between them:

    1

    Source rows

    What came out of Oracle EBS (counts supplied with the dataset).

    2

    Staged rows

    What passed load-readiness and was transformed into the Fusion shape.

    3

    Fusion rows

    What actually landed in Oracle Fusion, with success and failure counts from the load logs.

    On top of the counts, every data element was reconciled at row and column level — so the check was not merely “did the totals match” but “did each field arrive intact.”

    An auditable trail, not a claim

    Load logs recorded success and failure counts per object. A shared error tracker carried every exception — with the resolution or workaround applied — from first observation to closure. Together with the per-object reconciliation packs, this gave PwC evidence to hand to the business and to auditors, without waiting on a manual reconciliation exercise.

    Per-object recon packs
    Source vs staged vs Fusion counts for each object.
    Load logs
    Per-object success/failure counts from every run.
    Error tracker
    Each exception logged with its resolution or workaround.

    Frequently asked questions

    What does “row- and column-level reconciliation” mean?

    Beyond matching total row counts per object, each individual field (column) is validated against the source, so both completeness and content are proven — not just totals.

    What is the difference between load-readiness and business validation?

    Load-readiness (Syntra) checks whether a record will pass Fusion’s technical import rules; business validation (PwC) checks whether the data is business-correct. Separating them gives every reject a clear owner.

    What evidence was produced?

    Per-object reconciliation packs (source → staged → Fusion), load logs with success/failure counts, and a shared error tracker recording each exception and its resolution.

    Planning an Oracle EBS to Oracle Fusion migration?

    Tell us your source EBS footprint, modules and business units — we'll scope a mock-run-based transform-and-load plan on the Syntra ETL platform.