TotalApp Docs

Financial Audit & Fintech Ops — Overview

What the add-on is, its 4 sub-groups, who uses it, and the two-stage hybrid AI engine pattern used across the module.

Module Overview

The Financial Audit & Fintech Ops add-on is a wired, purchasable package gated by a tenant entitlement and per-user FinAudit roles. It covers the full internal-audit and fintech-operations lifecycle for a finance-heavy organisation: configuring jurisdiction-aware audit rules and AML/sanctions screening, reconciling transactions across bank/gateway/POS/ERP channels and flagging fraud, monitoring treasury liquidity, FX exposure and credit risk, and executing payments clearing, chargeback resolution and escrow lifecycles.

At a glance

  • Domain: FINAUDIT
  • URL prefix: /finaudit/…
  • Governance screens: Roles, Audit & Fintech Ops Team (staff)
  • 16 operational screens across 4 sub-groups (see below)
  • Default roles: Chief Audit & Fintech Ops Manager, Internal Auditor, Compliance & AML Officer, Treasury Analyst, Payments Operations Specialist

The 4 sub-groups

The add-on’s 16 operational screens are organised into four titled sub-groups, on top of the standard Administration & Access group (Roles, Staff).

GroupScreens
Audit Rules & Compliance SetupAudit Rules Registry, AML & Sanctions Config, Chart of Accounts Audit, Regulatory Reporting Setup
Transaction Audit & ReconciliationInvoice Matching Audit, Multi-Source Audit & Reconciliation (Bank Reconciliation), Fraud Detection & Anomaly Audit, Ledger Variance Audit
Treasury & Liquidity OperationsCash Flow Forecast, FX Exposure & Risk, Credit Risk & Limits, Liquidity Buffer
Payments & Clearing ExecutionPayout Routing Rules, Chargeback & Dispute, Settlement & Clearing, Financial Escrow

See Audit Rules & Compliance Setup, Transaction Audit & Reconciliation, Treasury & Liquidity Operations and Payments & Clearing Execution for the per-screen breakdown of each group.

Who uses it

Financial Audit & Fintech Ops targets internal auditors, compliance & AML officers, treasury analysts and payments operations specialists — the people who decide audit findings, compliance posture, treasury risk exposure, and payment operations health. A tenant activates the finaudit package, then assigns each team member one of the five default roles (or a custom role built from the same capability catalog) as a domain “hat” — see Roles & Capabilities.

Key concepts

Jurisdiction-aware audit rules

Audit Rules Registry, AML & Sanctions Config, Chart of Accounts Audit and Regulatory Reporting Setup automatically adapt their form fields and audit logic to the selected jurisdiction (Turkey/US/Default) — the same category of audit produces a different set of fields and checks depending on which jurisdiction is selected.

AML & sanctions screening

AML & Sanctions Config is the compliance layer that governs what counts as a sanctionable counterparty or suspicious pattern before any transaction-level audit runs against it.

Multi-source reconciliation

Invoice Matching, Bank Reconciliation, Fraud/Anomaly Detection and Ledger Variance Audit all reconcile records across independent channels — bank, payment gateway, POS, ERP internal — each with its own dedicated AI analysis.

Treasury risk trio

Cash Flow Forecast, FX Exposure & Risk and Credit Risk & Limits monitor forward-looking liquidity, currency and counterparty risk; Liquidity Buffer tracks the Basel III–style capital cushion behind them.

Payments clearing trio

Payout Routing Rules, Chargeback & Dispute and Settlement & Clearing run the operational side of moving money; Financial Escrow holds funds against a release condition or milestone.

Two screen generations

The 4 Audit Rules & Compliance Setup screens automatically adapt their fields to the selected jurisdiction; the other 12 screens run a dedicated AI analysis per record type — both go through the same AI-engine-aware second stage (see below).

The two-stage hybrid engine

Every screen in this add-on that does AI work follows the same two-stage pattern, and it’s worth understanding once rather than per-screen:

Stage 1 — Embedding Engine
Stage 2 — AI Synthesis
Result saved to ledger
StageWhat it doesWhere it runs
Stage 1 — Embedding Engine11 semantic-search helpers — one per record type (ledger entries, transactions, invoices, treasury records, FX hedges, counterparties, regulatory documents, routing rules, dispute evidence, settlement batches, escrow accounts) — all built on the same shared Finding Similar Records logic used across the app. It never throws an error; if it can’t complete a search it degrades gracefully and simply returns no matches.Always the user’s own local Ollama instance — this stage never proxies through the TotalApp server.
Stage 2 — AI Synthesis LayerThe older 4 screens (Audit Rules & Compliance Setup) automatically adapt their fields and logic to the selected jurisdiction (Turkey/US/Default). The newer 12 screens each run their own dedicated AI analysis (bank reconciliation, fraud/anomaly detection, invoice matching, ledger variance, cash flow forecast, FX exposure, credit risk, liquidity buffer, payout routing, chargeback/dispute, settlement/clearing, financial escrow), all following the identical AI-engine-routing pattern.Routes through whichever engine the user picked in Settings → Agentic — Ollama/local-LLM/web-LLM run entirely in the browser and never touch the server; local-CLI/API route to the server, which calls the Anthropic or Cohere API depending on the selected mode.

Embedding search is not wired into the UI yet

Unlike Patent & R&D Operations and Customs & Global Trade Operations, where Finding Similar Records is a mandatory first UI step (e.g. a “Find Similar Patents” or “Search Similar Rules” button), the 11 FinAudit semantic-search helpers exist purely as background infrastructure today — none of the 16 screens use them yet. Users go directly from filling the form to clicking “AI Analyze” (Stage 2); there is no “find similar records first” step anywhere in this add-on. This is infrastructure ready for a future UX, not a flaw — the underlying search logic is tested and working, just not yet surfaced on any screen.

Every server call is logged (duration, model name, success/fail, screen name) — visible in Settings → Analytics. When Ollama/local-LLM/web-LLM is selected, Stage 2 runs entirely in the browser and never hits the server at all — the Render server has no network path to a user’s local Ollama.

Frequently Asked Questions

What do I need to use the Financial Audit & Fintech Ops add-on?
The tenant must have the finaudit package activated, and your user must hold a FINAUDIT domain role. See Roles & Capabilities.
Do I need to click “Find Similar” before running AI Analyze?
No. Unlike Patent & R&D Operations and Customs & Global Trade Operations, this add-on’s 11 Finding Similar Records functions are not wired into any screen’s UI yet — you go straight from the form to “AI Analyze”. See Screen Workflows.
Do I need Ollama installed to use this add-on?
Only if your Settings → Agentic engine is set to Ollama, Local LLM, or Web LLM — those route entirely on your machine/browser. With Local CLI or API, the server handles Stage 2 instead.
Are external systems like real PSPs, a general ledger, or bank rails integrated?
No. Most action buttons (adjust variance, freeze account, trigger sweep, create FX hedge, trigger routing failover, submit dispute evidence) only update the screen’s own record — there is no real downstream integration. See Screen Workflows for the exact list of what is and isn’t a live integration.
Can a Financial Audit & Fintech Ops user see other add-ons’ data?
No. A FINAUDIT role exposes only FINAUDIT and GLOBAL capabilities.
Where is this add-on’s data stored?
Under the tenant’s finaudit-ops/ module folder on the server — one JSON file per screen (e.g. finaudit-audit-rules.json, finaudit-bank-reconciliations.json), plus roles/roles.json for role definitions. Curated seed copies live in server/seed/tenants/default/finaudit-ops/ for the admin Seed Data Transfer tool.
Which action buttons are genuine record updates rather than placeholders?
recalibrateCreditLimit (Credit Risk) and reallocateLiquidityBuffer (Liquidity Buffer) directly update their own record field and are not placeholders. See Screen Workflows.