Audit-grade epic systems data validation across three reconciliation tiers — Clarity vs Cogito, Fusion staging vs Clarity source, Fusion posted vs Clarity at load instant. Signed evidence pack per fiscal period for SOX, HIPAA and Joint Commission.
Errors can be introduced at extraction, transformation or load. Validation has to catch all three — or the first month-end close after go-live surfaces them at the worst possible moment.
Most ERP migration validation strategies reconcile at one tier — usually source vs target row counts and sum totals after load. That is necessary but not sufficient for Epic to Fusion. Clarity is a SQL Server mirror of Chronicles with a configurable ETL refresh lag (typically 2 hours operational, 24 hours analytical). Cogito is the analytics layer that often pulls a different freshness slice. The legacy ERP being retired has its own period-close cadence. The Fusion target has FBDI staging tables that can lose rows in transform errors. A one-tier validation misses errors that surface in production three months later.
Syntra ETL's epic systems data validation runs three tiers per load. Tier 1: Clarity source row-count + sum-check vs Cogito reference row-count + sum-check (catches Clarity ETL lag and partition gaps). Tier 2: Fusion staging row-count + sum-check vs Clarity source post-transform (catches mapping errors and FBDI template validation failures). Tier 3: Fusion posted row-count + sum-check vs Clarity at-the-load-instant snapshot (catches load failures, rejected rows and ESS job errors). Each tier produces a signed reconciliation artefact.
The validation harness is fully automated — runs as part of every FBDI Journal Import, every HDL Worker load, every Inventory Transaction Import. Within hours of each load, the three-tier reconciliation result is available. Variance beyond tolerance halts the load promotion. For a regional health system in parallel run, the validation cycle produces a signed monthly evidence pack 5–7 business days after period close. Two consecutive clean months are the cutover gate.
Validation harnesses pre-built for the highest-stakes downstream domains. Configure thresholds, the harness runs automatically.
Fusion GL AR balance per service area reconciled to Resolute AR aging at the same instant. Zero variance threshold. Cross-checked against Cogito period-end AR snapshot.
Fusion GL trial balance per legal entity reconciled against Cogito period-end snapshot and against the legacy ERP at the same period close. To the cent.
Every active CLARITY_EMP and CLARITY_SER record reconciled to a Fusion Worker. Provider DFFs (NPI, DEA, specialty) preserved. Zero duplicate threshold.
On-hand quantity + 340B-aware cost + transaction count reconciled per location per item. Cross-checked against Cogito pharmacy snapshot where available.
Surgical case-cart material consumption reconciled per service line. Implant tracking + lot + expiration preserved through migration.
Lab reagent supplier master + contract pricing reconciled. Instrument-linked materials catalogue cross-checked against Beaker's active workflows.
A repeatable validation cadence aligned to month-end close. Automated reconciliation, controller review, signed evidence pack.
Every overnight delta load (Clarity → Fusion incremental refresh) runs through three-tier reconciliation. Results posted to operations dashboard within 2 hours. Variances triaged daily by the migration ops team.
Resolute AR aging, Cogito period-end snapshot, Fusion GL trial balance captured at the same instant. Stored hash-signed for the evidence pack.
Clarity source vs Cogito reference reconciliation per partition per service area. Variance log produced. Any variance beyond tolerance triaged by Clarity analysts.
Fusion staging vs Clarity source post-transform reconciliation per FBDI template. Mapping errors caught here. Triage by mapping engineers and Resolute admin.
Fusion posted vs Clarity at-load-instant reconciliation per ESS job. Load failures and rejected rows caught here. Remediation plan produced for any variance.
Full evidence pack assembled, signed by controller and internal audit, archived for SOX/HIPAA/Joint Commission retrieval. Two consecutive clean monthly packs are the cutover gate.
Six concrete regulatory contexts where the epic systems data validation evidence pack is the difference between passing audit and writing checks.
Tier 3 reconciliation evidence demonstrates effective ITGC controls over financial reporting. External auditors retrieve the evidence pack directly during their walkthrough.
Joint Commission surveyors expect to see audit trails for any data system change affecting clinical or billing operations. The signed evidence pack satisfies on first-pass retrieval.
Audit control standard requires demonstrable logging of every PHI access. The actor log per load is HIPAA-grade — actor, timestamp, source hash, target hash.
CMS cost report (typically May 31 federal) depends on accurate AR aging and supply consumption history. The reconciliation evidence pack feeds cost-report preparation directly.
Non-profit 990 reporting depends on consolidated financials per legal entity. Trial-balance reconciliation evidence is included in 990 prep workpapers.
US HHS OIG audits (340B, billing compliance, etc.) require demonstrable data lineage. The hash-signed evidence chain demonstrates lineage from Clarity through Fusion.
Epic systems data validation is the systematic post-load reconciliation that confirms every Clarity-extracted row landed correctly in Oracle Fusion — at row, sum and hash level — before the legacy ERP is cut to read-only. For Resolute HB/PB AR feeds, validation reconciles Fusion GL trial balance against Resolute AR aging and against the Cogito period-end snapshot to the cent. For CLARITY_EMP+SER → HDL Worker, validation confirms every active employee and active provider in Clarity exists as a Fusion Worker with correct role and DFFs. For Willow/OpTime/Beaker → Inventory, validation confirms on-hand quantity, cost and transaction count match Clarity at the snapshot timestamp. Output: a signed evidence pack per fiscal period.
Because errors can be introduced at three different stages, and each tier catches a different class of error. Tier 1 (Clarity source vs Cogito reference) catches Clarity ETL lag — Clarity is a SQL Server mirror with a configurable refresh lag versus Chronicles, and Cogito sometimes pulls a different freshness slice. Tier 2 (Fusion staging vs Clarity source post-transform) catches mapping errors — a Resolute posting rule mis-mapped to the wrong GL account would show up here. Tier 3 (Fusion posted vs Clarity at the load instant) catches load failures — a Fusion ESS job error, a row rejection, a journal that posted to a wrong period. Without all three tiers, errors slip through and surface during the first month-end close after go-live.
Cogito is Epic's analytics layer — built on top of Clarity but with its own reporting models and refresh cadence. For most downstream finance/HCM/SCM data, Cogito provides an independent count and sum reference that confirms Clarity's freshness and integrity at a given point in time. The epic systems data validation harness queries Cogito for row counts and sum totals per period per service area and compares against Clarity. Variance beyond tolerance flags the partition for analyst review. Cogito also serves as the source for trial-balance and period-end snapshots that the Fusion GL reconciles against — the same numbers that historically anchored your CFO sign-off, used to anchor the migration sign-off.
A signed evidence pack per fiscal period — typically per month for a parallel-run window. Contents: (1) Reconciliation summary at the top — Clarity vs Cogito vs Fusion to the cent per service area and account class; (2) Tier 1 detail — Clarity vs Cogito row counts and sum checks per partition; (3) Tier 2 detail — Fusion staging vs Clarity post-transform; (4) Tier 3 detail — Fusion posted vs Clarity at load instant, with FBDI ESS job IDs; (5) Hash-signed file inventory of every extract, transform output and load payload; (6) Actor log of every service-account read for HIPAA chain of custody; (7) Variance log with disposition for any non-zero reconciliation differences. Format: PDF + machine-readable JSON for SOX and HIPAA auditor retrieval.
Resolute HB/PB AR aging is the highest-stakes validation in any Epic to Fusion migration because it's the number the CFO looks at first on day one. Validation strategy: Resolute generates the AR aging directly (the system of record stays Resolute) and the consolidated balance posts to Fusion GL via the downstream journal feed. The epic systems data validation reconciles Fusion GL AR balance per service area against Resolute AR aging at the same instant. They must agree to the cent. The validation also confirms the period-end Cogito AR snapshot agrees with Resolute AR aging — that catches Clarity ETL lag. Variance threshold is zero. Any non-zero variance halts the cutover until disposed.
Three checks. First, every active employee in CLARITY_EMP at the cutover snapshot must exist as a Fusion Worker with employee role set and correct DFFs. Second, every active provider in CLARITY_SER at the cutover snapshot must exist as a Fusion Worker with provider role set and provider DFFs (NPI, DEA, specialty, board cert) preserved. Third, every person who appears in both tables (employee + provider) must exist as one and only one Fusion Worker with both roles set — no duplicates. Validation runs an automated cross-check and produces a variance report. Threshold is zero variance. Typical issues caught: employee-provider name mismatches that prevented merge, retired providers that were filtered incorrectly, contracted physicians without employee record.
Inventory validation reconciles three things per location per item. First, on-hand quantity: Fusion Inventory on-hand at the cutover snapshot must agree with Willow/OpTime/Beaker on-hand at the same instant. Second, inventory value (cost): Fusion Inventory cost basis must agree with Willow/OpTime/Beaker cost (note Willow uses 340B-aware costing — Fusion has to mirror). Third, transaction history: total transaction count and net transaction value per item per period must agree between Clarity-extracted history and Fusion-loaded history. Variance threshold is item-level zero quantity, $50 cost variance per item (typically rounded due to cost averaging differences across systems). Cross-validated against Cogito period-end inventory snapshot where available.
Validation runs automatically as part of every load and produces results within hours. The full evidence pack per fiscal period takes 1–2 days to assemble and sign — most of that is the controller and internal audit review, not the technical reconciliation. For a regional health system in 1–2 month-end close parallel run, the validation cycle is: daily delta reconciliation runs nightly with sub-2-hour results; weekly variance review with the controller; monthly evidence pack assembled within 2 days of period close, reviewed and signed within 5 business days. Cutover sign-off requires two consecutive clean monthly evidence packs — typical timeline: 8–10 weeks from go-live preview to legacy ERP read-only cut.
Book a 30-minute reconciliation deep-dive. We'll walk through your Resolute AR validation, Cogito snapshot strategy, three-tier evidence pack design, and timeline to cutover. Concrete validation plan before the call ends.