Excel-based cohort logic, version drift, auditor frustration, regulator scrutiny.
The scenario
Who: A 25-adviser AFSL practice running a multi-year fee-for-no-service remediation across its adviser book.
The problem: Cohort definitions shift as edge cases surface; the licensee asks why the provision moved between reviews; the team can’t reconstruct the calculation from three months ago.
Cohort logic versioned and auditable, every figure traces to platform fee rows, and the provision reconciles to the dollar across Xplan, HUB24, Netwealth and BT Panorama.
What it produces
ASIC’s consumer remediation framework is set by Regulatory Guide 277 (updated September 2022). RG 277 expects firms to identify affected cohorts, quantify impact with reproducible methodology, and demonstrate that the remediation has reached every consumer it should. Fydis identifies affected cohorts from operational data, quantifies impact with cohort-level lineage, captures two-eyes sign-off, and produces the audit pack with reproducible methodology versioning. The numbers reproduce from the same inputs.
“Q3 remediation provision, fee-for-no-service review across the adviser book?”
$2,451,392
RetrievePlatform fee ledgers · row-level lineage
VerifyXplan · HUB24 · Netwealth · BT Panorama
EvidenceResponsible Manager · 2 sign-offs
EvidenceAudit pack · PDF + JSON · reproducible
The analysis, step by step
- 01RetrieveCohort rule defined and versioned (e.g. fee rows with no documented service, 2014-2021)
- 02VerifySource rows resolved with content hash across the platform fee ledgers
- 03VerifyImpact quantification function applied (versioned, library-managed)
- 04EvidenceTwo-eyes sign-off, Responsible Manager and head of advice
- 05EvidenceRemediation submission generated · audit pack PDF + JSON
Frequently asked
What if the cohort definition changes mid-program?
Each version is preserved. Old figures retain their original hash; new versions are new artefacts. Reconciliation is automatic.
How is this different from a remediation tool?
Your remediation tool’s figures run through Fydis. We don’t replace the calculation engine; we make every figure auditable.
Where does ASIC REP 798 fit in?
REP 798 is ASIC’s 2024 AI governance report, separate to the remediation framework. RG 277 is the consumer remediation guide. They overlap when the remediation calculation is AI-assisted: REP 798 then signals the governance bar, RG 277 sets the remediation bar. Fydis produces the evidence chain for both.
Run this on your data.
The live demo runs this exact analysis on sandboxed data. Book a 30-minute briefing and we’ll show you the same chain on your data, your approval policy, and your regulator clause map.