Continuous Finance validation script

From individual AI use cases to Continuous Finance Decision Runtime

The FP&A technology direction changes the validation target. The strongest story is no longer invoice, journal, contract, credit, or spend in isolation. The stronger proof is that Sphere connects Continuous Accounting, Continuous Close, Driver-Based Planning, Daily Actuals, Automated Forecasting, Intercompany Automation, Automated Controls, and Variance Intelligence into one governed decision system.

Open platform services Open scenario JSON

End-to-end story validation should see

Business EventProduction, invoice, maintenance, FX, IC, journal, or actuals event occurs.

Driver ChangesDriver Ontology resolves production, price, volume, utilization, headcount, capex, or maintenance impact.

Forecast ChangesForecast Intelligence refreshes revenue, cash, EBITDA, margin, and confidence.

Accounting ImpactFinancial Event Fabric and Continuous Close Engine identify accrual, journal, reconciliation, or IC impact.

Control ValidationPolicy Runtime validates materiality, authority, SoD, evidence, and embedded controls.

Journal AutomationBlackLine/SAP path drafts or validates journal and reconciliation action.

Close Readiness UpdateContinuous Close recalculates readiness, blockers, owner, and critical path.

CFO InsightVariance/Narrative agents explain what changed, why, what happens next, and what to do.

Agent inventory required for validation

Agent	Priority	Demo	Purpose
Financial Event Agent	P1	Yes	Converts operational events into financial signals and impact triggers.
Driver Intelligence Agent	P1	Yes	Maps operational driver movement to financial outcomes.
Forecast Agent	P1	Yes	Refreshes forecast, confidence, and financial impact.
Variance Agent	P1	Yes	Explains plan/forecast/actual variance with driver attribution.
Journal Agent	P1	Yes	Scores journal risk and recommends approval or escalation.
Accrual Agent	P1	Yes	Finds missing accruals and drafts evidence-backed recommendations.
Intercompany Agent	P1	Yes	Matches, explains, and recommends IC resolution.
Evidence Agent	P1	Yes	Assembles evidence packs, lineage, and replay proof.
Close Agent	P1	Yes	Computes close readiness and critical path blockers.
Policy Agent	P1	Yes	Evaluates rules, authority, controls, and action boundaries.
Narrative Agent	P2	Yes	Produces management commentary from evidence and drivers.
Capitalization Agent	P1	Yes	Determines expense vs capitalize and AUC readiness.
AUC Monitoring Agent	P1	Yes	Monitors aged AUC, completion signals, contamination, and impairment risk.
Asset Completion Agent	P1	Yes	Detects technical completion and readiness to capitalize.
Fixed Asset Policy Agent	P1	Yes	Applies fixed asset policy, materiality, precedent, and SoX controls.
SoX Evidence Agent	P1	Yes	Assembles control evidence for capitalization and reporting decisions.
Allocation Driver Agent	P1	Yes	Validates allocation and recharge drivers.
PaPM Lineage Agent	P1	Yes	Explains formula, version, source driver, output, and posting target.
Recharge Variance Agent	P1	Yes	Explains bill-to-actual breaks and recommends adjustment or escalation.
Allocation Policy Agent	P1	Yes	Applies materiality, ownership, and approval controls to allocations.
Reporting Evidence Agent	P1	Yes	Builds evidence packs for reported numbers and commentary.
ICFR Control Agent	P1	Yes	Maps reporting assertions to controls, tests, and exceptions.
Narrative Assurance Agent	P1	Yes	Validates commentary against evidence and variance drivers.
Certification Routing Agent	P1	Yes	Routes reporting signoff and captures accountable approvers.
SAP Skill Orchestration Agent	P1	Yes	Treats SAP Joule/BTP/SAP-native actions as governed runtime skills.
BTP Extension Governance Agent	P1	Yes	Validates extension ownership, data boundary, and control impact.
SAP Coexistence Agent	P1	Yes	Resolves ECC, CFIN, S/4HANA, and PaPM mapping state.
SAP Action Control Agent	P1	Yes	Blocks unsafe SAP write-back and routes controlled execution.
Data Readiness Agent	P1	Yes	Scores finance dataset readiness for continuous finance decisions.
Lineage Agent	P1	Yes	Traces source system, transformation, version, timestamp, and owner.
Mapping Completeness Agent	P1	Yes	Finds missing HDS, OneData, Foundry, SAP, or PaPM mappings.
Freshness Guardrail Agent	P1	Yes	Reduces confidence or blocks action when data is stale.
PRA Agent	P1	Yes	Extracts and reconciles production revenue accounting evidence.
Pilot Readiness Agent	P1	Yes	Scores pilot readiness across data, controls, agents, evidence, and replay.
Pilot Evidence Agent	P1	Yes	Defines pilot proof requirements and builds evidence contracts.
Pilot Dependency Agent	P1	Yes	Finds missing source, owner, integration, and policy dependencies.
Pilot Value Agent	P1	Yes	Links pilot outcomes to close compression, automation, forecast, and control value.
Reconciliation Agent	P1	Yes	Explains breaks, aging, root cause, and owner.
Settlement Agent	P1	Yes	Recommends IC netting, settlement, or adjustment.

Readiness audit framework

Area	Question
Data	Do we have realistic data?
Ontology	Do concepts exist?
Events	Are business events modeled?
Policies	Are rules implemented?
Evidence	Can recommendations be proven?
Replay	Can decisions be replayed?
Agent Logic	Is reasoning implemented?
Human Workflow	Can users intervene?
Integration	Are SOR integrations demonstrated?
Controls	Are approvals enforced?
Learning	Can agents improve?
Observability	Can runtime be monitored?

Additional enterprise data required

Domain	Data required
FP&A	Historical forecasts, latest estimates, actuals, variance explanations, driver hierarchies.
Continuous Close	Close calendar, close tasks, reconciliation exceptions, journal approvals.
Intercompany	IC balances, IC disputes, ownership, settlement history, counterparty mappings.
BlackLine	Journal analyzer outputs, matching data, reconciliation data, close status.
Daily Actuals	Lighthouse Daily Actuals, UDP datasets, Databricks models, SAP actuals, Treasury balances.
PRA / Evidence	Operator statements, PDFs, spreadsheets, production data, SAP records, calculation support.
Fixed Assets / AUC	AUC balances, project/WBS data, work orders, technical completion events, asset records, SoX controls.
PaPM / Allocations	PaPM formulas, allocation drivers, bill-to-actual records, SAC outputs, receiving entity mappings.
Reporting / Workiva	Reported metrics, reporting package data, ICFR controls, approvals, commentary, source actuals.
SAP BTP / Joule	SAP skill catalog, BTP extension registry, ECC/CFIN/S4 mappings, PaPM objects, ServiceNow handoffs.
Finance Data Readiness	Quantum HDS mappings, Lighthouse Daily Actuals, UDP datasets, Databricks tables, OneData definitions, Foundry mappings.
Continuous Accounting Pilots	Drilling cost AGT/GOA, ANZ region, P&L-impact-led, terminal data automation, accrual automation, aviation pilot scope and owners.

Must-have demonstration scenario · readiness 4.5/5

Continuous Close Command Center

Audience: CFO, Controller, FP&A Lead
Story: Close status moves from unknown until period-end to visible every day.

Close Agent · Accrual Agent · Journal Agent SAP ECC · SAP S/4HANA · BlackLine

Demo element	Detail
Demo	Show 92% close readiness, then drill into 3 accrual issues, 2 intercompany mismatches, and 1 journal awaiting approval with evidence, policy, action, impact, and replay.
Data required	SAP journal data, BlackLine reconciliations, Close calendar, Close tasks, Approvals
Integrations	SAP ECC, SAP S/4HANA, BlackLine, ServiceNow
Agents	Close Agent, Accrual Agent, Journal Agent, Policy Agent, Evidence Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	4	Do we have realistic data?
Ontology	5	Do concepts exist?
Events	4	Are business events modeled?
Policies	5	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	4	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	4	Can agents improve?
Observability	5	Can runtime be monitored?

Must-have demonstration scenario · readiness 3.7/5

Driver-Based Forecasting

Audience: FP&A Lead, CFO, Business Finance
Story: Business changes production assumptions and the forecast updates with financial impact.

Driver Intelligence Agent · Forecast Agent · Variance Agent UDP · Databricks · SAP

Demo element	Detail
Demo	Move Oil Production -5% and immediately show revenue, cash, EBITDA, and margin changes with an explanation drawer.
Data required	Production drivers, Volumes, Prices, Historical forecast, Actuals
Integrations	UDP, Databricks, SAP, Foundry
Agents	Driver Intelligence Agent, Forecast Agent, Variance Agent, Narrative Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	4	Can recommendations be proven?
Replay	4	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	3	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	4	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 3.9/5

Automated Variance Intelligence

Audience: FP&A Lead, Segment Finance, CFO
Story: Variance explains itself instead of relying on manual commentary assembly.

Variance Agent · Root Cause Agent · Action Agent SAP · UDP · Databricks

Demo element	Detail
Demo	Open revenue variance -$120M and show attribution: 60% volume, 25% price, 10% FX, 5% timing, followed by recommended actions.
Data required	Actuals, Forecast, Drivers, FX, Commodity prices
Integrations	SAP, UDP, Databricks, Market data
Agents	Variance Agent, Root Cause Agent, Action Agent, Narrative Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	4	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	4	Can recommendations be proven?
Replay	4	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	3	Can users intervene?
Integration	4	Are SOR integrations demonstrated?
Controls	4	Are approvals enforced?
Learning	4	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.1/5

Real-Time Intercompany Resolution

Audience: Controller, R2R Lead, Project LIBRA Team
Story: Intercompany mismatch is detected, explained, routed, and resolved before close pressure builds.

Intercompany Match Agent · Root Cause Agent · Settlement Agent SAP · BlackLine · Treasury

Demo element	Detail
Demo	Show Company A $8.2M payable vs Company B $7.7M receivable; agent identifies timing issue, payment pending, and recommends resolution.
Data required	IC balances, IC disputes, Settlement history, Counterparty mappings
Integrations	SAP, BlackLine, Treasury
Agents	Intercompany Match Agent, Root Cause Agent, Settlement Agent, Evidence Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	4	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.5/5

Automated Journal Intelligence

Audience: Controller, R2R Lead, Internal Controls
Story: Journal risk is checked against policy, materiality, prior history, SoD, and evidence before approval or escalation.

Journal Intelligence Agent · Policy Agent · Risk Agent SAP · BlackLine · Microsoft Entra ID

Demo element	Detail
Demo	Open a journal, show risk score, policy evaluation, BlackLine/SAP state, and approve-or-escalate recommendation.
Data required	Journal history, Policies, Approvals, SAP journals, BlackLine journal analyzer outputs
Integrations	SAP, BlackLine, Microsoft Entra ID
Agents	Journal Intelligence Agent, Policy Agent, Risk Agent, Evidence Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	4	Do we have realistic data?
Ontology	5	Do concepts exist?
Events	4	Are business events modeled?
Policies	5	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	4	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	4	Can agents improve?
Observability	5	Can runtime be monitored?

Must-have demonstration scenario · readiness 3.8/5

Daily Actuals / Finance Pulse

Audience: CFO, FP&A Lead, Controller
Story: Current financial position is continuously visible rather than discovered after month-end.

Daily Actuals Agent · Financial Event Agent · Variance Agent SAP · UDP · Databricks

Demo element	Detail
Demo	Show today revenue, margin, cash, EBITDA, and forecast drift; ask what changed today and trace source freshness.
Data required	Lighthouse daily actuals, UDP datasets, Databricks models, Treasury data, SAP actuals
Integrations	SAP, UDP, Databricks, Treasury
Agents	Daily Actuals Agent, Financial Event Agent, Variance Agent, Forecast Refresh Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	5	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	4	Can recommendations be proven?
Replay	4	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	3	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	4	Are approvals enforced?
Learning	3	Can agents improve?
Observability	5	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.2/5

Continuous Accrual Automation

Audience: Controller, Operations Finance, Internal Controls
Story: Maintenance completed but invoice not received becomes a recommended accrual with evidence and policy.

Accrual Agent · Evidence Agent · Policy Agent SAP · Ariba · Maintenance systems

Demo element	Detail
Demo	Show maintenance event, PO/receipt/history evidence, recommended accrual, policy validation, control validation, and SAP/BlackLine action path.
Data required	PO, Receipt, Maintenance events, Historical costs, SAP actuals
Integrations	SAP, Ariba, Maintenance systems, BlackLine
Agents	Accrual Agent, Evidence Agent, Policy Agent, Control Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	5	Are business events modeled?
Policies	5	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 3.6/5

PRA Evidence Intelligence

Audience: Upstream Finance, Controller, Audit
Story: PDFs, operator statements, spreadsheets, production data, and SAP records are transformed into evidence-linked reconciled finance decisions.

OCR Agent · Evidence Agent · Reconciliation Agent SAP · SharePoint/OpenText · Databricks

Demo element	Detail
Demo	Extract and reconcile PDF/operator statement/spreadsheet/SAP data, then show evidence lineage, confidence, calculation, and replay.
Data required	Operator statements, Production data, PDFs, Spreadsheets, SAP records
Integrations	SAP, SharePoint/OpenText, Databricks, Evidence Intelligence Runtime
Agents	OCR Agent, Evidence Agent, Reconciliation Agent, Calculation Agent, Policy Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	2	Do we have realistic data?
Ontology	3	Do concepts exist?
Events	3	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	3	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	4	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.1/5

Capitalization and AUC Intelligence

Audience: Controller, Fixed Assets Lead, Project Finance, Internal Controls
Story: Auto-capitalization and AUC decisions become policy-backed, evidence-linked, SoX-aware, and visible before close.

Capitalization Agent · AUC Monitoring Agent · Asset Completion Agent SAP Asset Accounting · SAP BTP · Project systems

Demo element	Detail
Demo	Open an AUC project, show technical completion, policy treatment, capitalization evidence, SoX impact, SAP asset path, and replay.
Data required	AUC balances, Project/WBS data, Work orders, Technical completion events, Fixed asset policy, SAP asset records
Integrations	SAP Asset Accounting, SAP BTP, Project systems, Evidence Intelligence Runtime
Agents	Capitalization Agent, AUC Monitoring Agent, Asset Completion Agent, Fixed Asset Policy Agent, SoX Evidence Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	5	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 3.6/5

PaPM Allocation and Recharge Automation

Audience: FP&A Lead, Controller, Allocation Process Owner
Story: Allocations and recharges become driver-linked, calculation-lineage-backed, policy-controlled, and replayable.

Allocation Driver Agent · PaPM Lineage Agent · Recharge Variance Agent SAP PaPM · SAP SAC · SAP

Demo element	Detail
Demo	Open a bill-to-actual variance, show PaPM formula version, source drivers, recharge recipient, policy threshold, recommended adjustment, and replay.
Data required	PaPM allocation outputs, Driver sources, Bill-to-actual records, SAC reports, Receiving entity mappings
Integrations	SAP PaPM, SAP SAC, SAP, Databricks, UDP
Agents	Allocation Driver Agent, PaPM Lineage Agent, Recharge Variance Agent, Allocation Policy Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	2	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	3	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	4	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	3	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	4	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 3.9/5

Statutory Reporting Evidence

Audience: Group Reporting, Controller, CFO, Internal Controls
Story: Reported numbers and commentary become evidence-backed, control-mapped, approved, and replayable.

Reporting Evidence Agent · ICFR Control Agent · Narrative Assurance Agent Workiva · SAP · BlackLine

Demo element	Detail
Demo	Open a board-pack metric, show source lineage, narrative support, ICFR control, owner certification, Workiva-style evidence package, and replay.
Data required	Reported metrics, Workiva/reporting package data, ICFR controls, Approvals, Variance commentary, Source actuals
Integrations	Workiva, SAP, BlackLine, UDP, Evidence Intelligence Runtime
Agents	Reporting Evidence Agent, ICFR Control Agent, Narrative Assurance Agent, Certification Routing Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	2	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	3	Are business events modeled?
Policies	5	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	3	Is reasoning implemented?
Human Workflow	5	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.2/5

SAP BTP and Joule Coexistence

Audience: SAP Architecture, AI Platform, Security, Finance Technology
Story: SAP-native capabilities participate as governed skills while the runtime coordinates cross-platform policy, evidence, replay, and accountability.

SAP Skill Orchestration Agent · BTP Extension Governance Agent · SAP Coexistence Agent SAP Joule · SAP BTP · SAP ECC

Demo element	Detail
Demo	Show a SAP-native validation skill, BTP extension, ECC/CFIN/S4 object mapping, action gateway policy check, optional SAP deep link, and replay.
Data required	SAP API/event catalog, BTP extension registry, ECC/CFIN/S4 mappings, PaPM objects, ServiceNow task/posting metadata
Integrations	SAP Joule, SAP BTP, SAP ECC, SAP CFIN, SAP S/4HANA, SAP PaPM, ServiceNow
Agents	SAP Skill Orchestration Agent, BTP Extension Governance Agent, SAP Coexistence Agent, SAP Action Control Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	5	Are rules implemented?
Evidence	4	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	4	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	5	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.0/5

Finance Data Readiness

Audience: Data Architecture, FP&A, Quantum/HDS Team, AI Platform
Story: Agents only recommend or execute when finance data freshness, lineage, mapping completeness, and source confidence are explicit.

Data Readiness Agent · Lineage Agent · Mapping Completeness Agent UDP · Databricks · OneData

Demo element	Detail
Demo	Open a Daily Actuals data product, show UDP/Databricks source freshness, Quantum HDS mapping, OneData/Foundry semantics, confidence guardrail, and blocked autonomous action for stale data.
Data required	Quantum HDS mappings, Lighthouse Daily Actuals, UDP datasets, Databricks tables, OneData definitions, Foundry ontology mappings
Integrations	UDP, Databricks, OneData, Foundry, SAP ECC, SAP CFIN, SAP S/4HANA
Agents	Data Readiness Agent, Lineage Agent, Mapping Completeness Agent, Freshness Guardrail Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	4	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	4	Can recommendations be proven?
Replay	4	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	3	Can users intervene?
Integration	4	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	5	Can runtime be monitored?

Must-have demonstration scenario · readiness 4.0/5

Continuous Accounting Pilot Tracker

Audience: FP&A Technology, Pilot Owners, Architecture, Controls
Story: The six FP&A pilot themes become inspectable delivery increments mapped to agents, source systems, policies, evidence, owners, readiness, and replay proof.

Pilot Readiness Agent · Pilot Evidence Agent · Pilot Dependency Agent SAP · BlackLine · SAP PaPM

Demo element	Detail
Demo	Open the six-pilot tracker, select Accrual Automation or Terminal Data Automation, and show agents, source data, evidence contract, policy gates, readiness score, and replay path.
Data required	Pilot scope, Owner mapping, Source-system catalog, Policy controls, Evidence requirements, Readiness scores
Integrations	SAP, BlackLine, SAP PaPM, UDP, Databricks, Terminal/operational systems
Agents	Pilot Readiness Agent, Pilot Evidence Agent, Pilot Dependency Agent, Pilot Value Agent

Open proof surface

Readiness audit

Area	Score	Question
Data	3	Do we have realistic data?
Ontology	4	Do concepts exist?
Events	4	Are business events modeled?
Policies	4	Are rules implemented?
Evidence	5	Can recommendations be proven?
Replay	5	Can decisions be replayed?
Agent Logic	4	Is reasoning implemented?
Human Workflow	4	Can users intervene?
Integration	3	Are SOR integrations demonstrated?
Controls	5	Are approvals enforced?
Learning	3	Can agents improve?
Observability	4	Can runtime be monitored?