page-11 · autonomous finance portfolio

Stakeholder Rebuttals
& Demo Pathway

Six architectural rebuttals to the hardest questions raised by CFO, Group Controller, CTO, CISO, External Auditor, and Transfer Pricing Counsel — followed by a scripted three-minute demonstration that shows the system working end-to-end.

Stakeholders 6 Functions

Demo Duration 3 Minutes

Demo Steps 6 (Step 0 = Setup, Steps 1–5 = Live)

Evidence Standard BigQuery · Immutable

Section 01 — Objection Handling

Stakeholder Rebuttals

Architectural responses to the six hardest questions raised across finance, technology, security, audit, and legal functions.

Stakeholder Concern Map — Who Owns What Layer

Stakeholder 01

CFO

Concern

"Our SAP implementation cost €4M. Why do we need an AI layer on top of it?"

Architectural Response

SAP is the system of record, not the intelligence layer. The AI sits above SAP, reads from it, and writes back only approved decisions with HITL records. It doesn't replace SAP — it makes the €4M investment productive by operating on what SAP already holds.

Page 04 Architecture — AI sits above ERP

Stakeholder 02

Group Controller

Concern

"I don't trust a model to flag IC mismatches. I've been doing this for 15 years."

Architectural Response

The model doesn't decide. It surfaces the anomaly with SHAP explanation and comparable prior mismatches. The correction cannot post to SAP without your explicit approval. Your 15 years of judgment remains the gate — the model is pattern recall at scale.

ADR-AF-01 — HITL write gate enforced

Stakeholder 03

CTO

Concern

"Three separate agent services is over-engineering. One service would be simpler."

Architectural Response

AP volume is 10× IC volume. Independent scaling, failure domains, and SLA management require service separation. A single service couples the latency profile and failure surface of three structurally different workloads — that's the operational risk, not a simplicity gain.

ADR-AF-04 — Independent scaling rationale

Stakeholder 04

CISO

Concern

"Banking API credentials and SAP credentials in the same GCP environment as the AI models is a security concern."

Architectural Response

VPC-SC perimeter, Secret Manager, and Workload Identity Federation enforce hard separation. No credential touches the same service account as model inference. The credential plane and the inference plane are architecturally isolated — not merely policy-isolated.

Page 04 — VPC-SC perimeter diagram

Stakeholder 05

External Auditor

Concern

"How do I verify that the SHAP explanation the Group Controller saw is the same one in the audit trail?"

Architectural Response

SHAP is written to BigQuery before the HITL checkpoint is created — same pattern as RevRec AI. The record is immutable, queryable, and the hitl_id is the join key between what the Controller saw and what BigQuery holds. One join. No ambiguity.

BigQuery · SHAP pre-write · one join

Stakeholder 06

Transfer Pricing Counsel

Concern

"Does the IC Reconciliation Agent create a documentation trail sufficient for OECD BEPS Action 13?"

Architectural Response

Every IC transaction processed generates an immutable BigQuery record with entity pair, amount, transfer price rate, agreement reference, and HITL approval. Queryable by tax counsel without a data engineer. Designed to meet OECD BEPS Action 13 country-by-country requirements structurally.

Page 07 — Compliance section · BEPS mapping

Rebuttal Architecture — How Each Concern Maps to a System Control

Section 02 — Scripted Demonstration

Demo Pathway

A scripted three-minute end-to-end demonstration. Six steps (Step 0 = setup, Steps 1–5 = live) from SAP mock to BigQuery audit trail — showing the IC Reconciliation Agent operating live.

Demo Sequence — Data Flow & System Transitions (3 minutes)

Demo Step 0 · 0:00 – 0:30

Setup — Veldtmann IC Transaction in SAP Mock

Show SAP S/4HANA mock environment — Veldtmann Group intercompany transaction between NL entity (Veldtmann BV) and DE entity (Veldtmann GmbH). Transaction: €2,340,000 management fee. Transfer price rate applied: 8.2%. Agreement reference: IC-MF-2024-003. Show both sides of the transaction — the receivable in NL, the payable in DE.

Say "This is what the system sees before any AI involvement — a raw intercompany transaction sitting in SAP. Transfer price applied, agreement referenced. The agent is about to read this and decide whether it's clean."

SAP S/4HANA Mock IC Dashboard BigQuery Cloud Run

Demo Step 1 · 0:30 – 1:00

Trigger — Month-End Pub/Sub Event Fires IC Agent

Show GCP Pub/Sub console — month-end-close topic. Publish message manually (simulating Cloud Scheduler trigger). Show IC Reconciliation Agent Cloud Run service receiving the message — status transitions from idle to active. Show subscription acknowledgement latency in the console.

Say "Month-end closes. A single Pub/Sub message triggers the IC agent — it picks up all open intercompany transactions from SAP in one batch. No cron job, no manual kickoff, no email to a team."

SAP S/4HANA Mock GCP Pub/Sub Cloud Run — IC Agent BigQuery

Demo Step 2 · 1:00 – 1:45 · KEY STEP

Inference — Extraction → Anomaly Score → SHAP → HITL Checkpoint Created

Show Cloud Run structured logs in real time. Walk through the four inference stages shown in the terminal: SAP extraction, anomaly scoring, SHAP value generation, HITL checkpoint creation in BigQuery. Each stage is a distinct log entry with a timestamp.

Say "Watch the log sequence. Extract from SAP. Score the anomaly — 0.87, above threshold. Generate SHAP values — those are written to BigQuery before the HITL checkpoint is created. The auditor's question is answered right here: the explanation is immutable by the time the Controller sees it."

Cloud Run — IC Reconciliation Agent · Structured Logs

2026-03-31T23:59:01.042Z INFO [ic-agent] Pub/Sub trigger received topic=month-end-close
2026-03-31T23:59:01.118Z INFO [ic-agent] SAP extraction started entity_pair=NL→DE period=2026-03
2026-03-31T23:59:01.843Z INFO [ic-agent] Transactions extracted count=47 open_mismatches=3
2026-03-31T23:59:02.011Z INFO [ic-agent] Anomaly scoring — transaction IC-NL-DE-20260331-0012
2026-03-31T23:59:02.284Z WARN [ic-agent] Anomaly score: 0.87 threshold=0.70 → FLAG
2026-03-31T23:59:02.301Z INFO [ic-agent] SHAP generation started model=lightgbm-ic-v3.2.1
2026-03-31T23:59:02.589Z INFO [ic-agent] SHAP values computed
transfer_price_rate_delta: +0.34
agreement_ref_mismatch: +0.28
counterparty_hist_pattern: +0.15
amount_vs_forecast: +0.10
2026-03-31T23:59:02.614Z INFO [ic-agent] SHAP written to BigQuery table=ic_shap_log row_id=shap-20260331-0012
2026-03-31T23:59:02.630Z INFO [ic-agent] HITL checkpoint created hitl_id=HITL-IC-01 status=PENDING_REVIEW
2026-03-31T23:59:02.631Z INFO [ic-agent] Group Controller queue notified controller@veldtmann.corp.internal

Cloud Run — IC Agent BigQuery · SHAP Log HITL Queue SAP S/4HANA Mock

Demo Step 3 · 1:45 – 2:30

HITL-IC-01 — Group Controller Reviews and Approves

Show Group Controller HITL dashboard — queue shows HITL-IC-01 pending review. Open the checkpoint: mismatch detail, SHAP waterfall chart (transfer price rate delta is the dominant factor at 0.34), five prior mismatches with same entity pair listed for comparison. Controller clicks Approve — status transitions to APPROVED, SAP write authorised.

Say "This is the gate. The correction cannot touch SAP without this approval. The Controller sees the SHAP explanation — which factor drove the flag — and five comparable prior mismatches so they have context. They approve. Now the write is authorised."

HITL-IC-01 · IC Reconciliation · Veldtmann NL → DE Pending Review

Transaction ID IC-NL-DE-20260331-0012

Amount € 2,340,000.00

Anomaly Score 0.87 — above threshold (0.70)

Top SHAP Factor transfer_price_rate_delta: +0.34

Agreement Reference IC-MF-2024-003 (mismatch with IC-MF-2025-001)

SHAP Record ✓ Written pre-HITL · shap-20260331-0012

SAP write blocked until approved

HITL Dashboard BigQuery · SHAP SAP Write Gate Cloud Run

Demo Step 4 · 2:30 – 2:45

SAP Write — Journal Correction Posts with hitl_id

Show SAP journal entry — correction posting with hitl_id=HITL-IC-01 embedded in the journal header text field. Show both the debit and credit legs of the correction. Timestamp matches the approval event from the HITL dashboard.

Say "The journal posts to SAP with the HITL ID embedded. The system of record now carries a traceable reference back to every piece of evidence the Controller reviewed. The AI wrote nothing to SAP without human approval."

SAP S/4HANA Journal Entry · hitl_id BigQuery Cloud Run

Demo Step 5 · 2:45 – 3:00

Audit Trail — BigQuery Full Evidence Package

Show BigQuery console — run the join query joining ic_transactions, ic_shap_log, and hitl_checkpoints on hitl_id. Result shows the full evidence package in a single result row: entity pair, amount, TP rate, SHAP values, agreement reference, Controller approval timestamp, and SAP journal reference.

Say "One query. No data engineer needed. The auditor can run this themselves. Entity pair, amount, transfer price rate, the SHAP values the Controller saw, the agreement reference, and the SAP journal number — everything OECD BEPS Action 13 requires, queryable on demand."

BigQuery — ic_audit_evidence · join on hitl_id

SELECT t.*, s.shap_values, h.approved_by, h.approved_at, h.sap_journal_ref
FROM `veldtmann.af.ic_transactions` t
JOIN `veldtmann.af.ic_shap_log` s USING (hitl_id)
JOIN `veldtmann.af.hitl_checkpoints` h USING (hitl_id)
WHERE h.hitl_id = 'HITL-IC-01';

-- Result (1 row, 14 columns, 28ms)

hitl_id	entity_pair	amount_eur	tp_rate	agreement_ref	top_shap_factor	approved_by	sap_journal_ref
HITL-IC-01	VLD-NL → VLD-DE	2,340,000	8.2% (exp 7.8%)	IC-MF-2024-003	tp_rate_delta +0.34	controller@veldtmann.corp.internal	SAP-JE-20260331-0441

BigQuery · 3-table join ic_shap_log hitl_checkpoints SAP S/4HANA

System Visibility Map — Which Systems Are Live at Each Demo Step

Three-minute demonstration — regulatory encoding visible at every step: Assuming classification as a high-risk AI system under EU AI Act Annex III, the SHAP-before-HITL sequence in Step 2 satisfies Art. 13 (transparency) and addresses the External Auditor's concern structurally. The write gate in Step 3 satisfies Art. 14 (human oversight). The BigQuery evidence package in Step 5 satisfies OECD BEPS Action 13 for Transfer Pricing Counsel — no data engineer required, queryable on demand. All six stakeholder rebuttals are observable live in the three-minute sequence.