Deterministic settlement with proof-gated disbursement

We don’t replace your rails—we are the release gate. Money moves only when the math replays identically and the right evidence is fresh and in-quorum.

Key Challenges?3-min-Demo ROI Calculator

Overview of VGoS as a release gate before rails/GL

ROI snapshot (baseline @ $250M GMV)

ROI definition:

~10%+

Steady-state profit lift (Yr-2)

$1,075,630

Year-1 net to client

$1,721,008

Year-2+ annual net

9,240 hrs

Automation hours saved / yr

After-Recovery ROI (vs Total Cost)

277%

Exact math

ROI_AR = Client(Y2+) ÷ [Verit(Y2+) + ResidualOps(Y2+)]
= $1,721,008 ÷ [$430,252 + $190,080]

Year-1 ROI (vs Total Cost)

85%

Recovery blending

Y1 client share @ 6 months = 50% (Client) / 50% (Verit)

ROI_Y1 = Client(Y1) ÷ [Verit(Y1) + ResidualOps(Y1)]
= $1,075,630 ÷ [$1,075,630 + $190,080]

Per-leg gross (annual)

• Leg 1 — Settlement: $727,740
• Leg 2 — Automation: $443,520
• Leg 3 — Spend Intel: $980,000

ResidualOps (Y2+): $190,080
Proof: 12 × 2,000 × 55% × 30% × $48

Leg 1 — Deterministic Settlement

• Hard recovery + proof-gated float ≈ $727,740 gross.
• Client share: 65% in Year-1, 80% in Year-2+.
• “Peace-of-mind yield”: smaller reserves, faster close, fewer firefights.

Leg 2 — AI Automation (Agents)

• ~9,240 hours/year across recon, disputes, audits, reruns, QA.
• You keep 100%* of the savings (enterprise license).
• Metered by reasoning consumed (work performed), not value skim.

Leg 3 — ML Spend Intelligence

• Invoice-verified savings across PSP, FX, Cloud, Fraud, Logistics.
• Economics: 95% client / 5% platform on realized savings.
• Year-1 realization: 9/12 months (months 4–12 after setup).

Client share (L1, Y2+)

80%

ML client share

95%

Go-live timeline

~90 days

GMV coverage

$50M–$1B

*Automation “100%” = client retains all automation value created. Agents are enterprise-licensed by reasoning consumed (work performed), not a share of value.

STRUCTURAL INSIGHT

Data → Proof → Trust → Capital

Together they form Verit’s thesis: Data becomes Proof → Proof becomes Trust → Trust becomes Capital.

1–4

Fix the math

→ truth at the transaction level

See steps

01 Input 02 Compute 03 Transcript 04 Digest

5–9

Fix the operations

→ trust at the organization level

See steps

05 Governance 06 Holds 07 Agents 08 Liquidity 09 Privacy

10–12

Fix the economy

→ value at the network level

See steps

10 VRC 11 Network 12 Economy

Beyond Savings: ROI Calculator →Explore the 12 steps

Where VGoS sits in your stack

Sources

ERP · PSP · Events · CSV · DWH · KYC/Tax

VGoS Gate

Replay = Digest ✓ · Acceptance ✓

Authorize(window_id){ALLOW[], HOLD[]}

Add to released rows: window_id, output_digest, provider_batch_id

Rails + GL

Stripe/Adyen/PayPal · ACH · BaaS · NetSuite/Tipalti

Common Sources

ERP (orders, returns)PSP exportsEvent bus (Kafka/Kinesis)Data warehouse (Snowflake/BigQuery)Fraud/KYC (vendor tokens)CSV drops (SFTP/S3/GCS)

Common Rails & Destinations

Stripe ConnectAdyen for PlatformsPayPal PayoutsTreasury/ACHBanking-as-a-ServiceNetSuite / GLTipaltiCase mgmt (Jira/ServiceNow)

10-second version

We gate disbursements right before your rails/GL. If Replay = Digest and required evidence is fresh and in-quorum, we authorize the ALLOW set; otherwise rows are HOLD with reason codes.

One API + 3 fields

Call Authorize(window_id) before release. Add window_id, output_digest, and provider_batch_id to released rows so everything reconciles by digest.

Your stack stays the same

Rails, ERPs, and your warehouse remain. VGoS is an overlay, not a rip-and-replace.

Three value legs — one compounding ROI

1) Deterministic Settlement

Kill leakage at the source (rounding/timing/duplicates/policy bugs) and unlock proof-gated float. Share model: 65% client in Year-1 → 80% in Year-2+.

2) AI Automation (Agents)

~9,056 hours/year across recon, disputes, audit prep, reruns, QA. You keep 100% of value*; agents are metered by work performed, not a value skim.

3) ML Spend Intelligence

Cohort benchmarks + invoice deltas on PSP, FX, Cloud, Fraud, Logistics. Savings are realized and shared 95%/5% (client/platform).

*“100%” refers to value retention: clients keep all automation value created. Agents are priced by reasoning consumed (work units).

Patent-backed architecture (why it’s different)

Our method is patent-pending (provisional filed). The design keeps settlement deterministic while improving control, auditability, and scale.

Deterministic windowing

Single writer + monotone watermark; same inputs → same outputs, every time.

Integer math & carry ledger

Caps/bonuses without drift. Fixed-precision accumulation; remainders assigned explicitly.

Replay-equal transcripts

Each window seals to a digest; replays must equal the digest before money moves.

Proof-gated disbursement

Freshness + Quorum over ACK/SPV/CT; missing evidence → deterministic HOLD with owners.

Versioned policy manifests

Save → canary → promote → rollback; every change is transcripted and scoped to a window.

Governance & recovery

Reason-coded holds, owner trails, and predictable audits; transcripted decisions.

Custom AI working for you

Add intelligence without risking determinism. Custom agents handle the busywork; custom models surface insights — proofs still gate disbursements.

Custom Agents

Stay on top of the process

From intake to release, custom agents collect and chase attestations, reconcile receipts, and only ping you when a decision is needed. They don’t change the math—they keep the process moving.

• Auto-collect missing docs; route to owners with SLAs
• Nudge on potential duplicates before HOLDs
• Watch Freshness & Quorum; summarize what’s blocking release

Licensing: You keep 100% of automation value*. Agents are priced by reasoning consumed(work units), not a value skim.

Custom Machine Learning

Understand payouts, spot drift

Custom models learn your programs to surface drift, anomalies, and policy insights—off the payout path, so settlement stays deterministic.

• Cohort & outlier detection across payees and programs
• Policy impact analysis before you ship a change
• Forecasts on holds, release timing, and cash needs

ML Spend Intelligence — re-rating that books real savings

Invoice-verified savings

We benchmark and re-rate PSP, FX, Cloud, Fraud, Logistics monthly. Actions are tied to invoice deltas—savings are realized, not predicted.

Aligned economics

Year-1 includes a 3-month setup ($50K). Thereafter, savings share is 95% client / 5% platform. Windowed realization is 9/12 months in Year-1.

Strategic leverage

Use realized savings to fund sharper pricing and promos, or to win bids. This offensive edge compounds as coverage grows.

Payments (PSP)

Margin deltas by tender/region; renegotiate with invoice proof.

Corridor markups vs mid; reduce volatility and leakage.

Cloud

Right-size compute/storage; commitment planning with usage evidence.

Fraud & Chargebacks

Tune thresholds for loss/min friction; track realized impact.

Policy: what must be true before we pay

Knobs

Bonus %, global/per-principal caps, required attestations (ACK/SPV/CT).

Outcomes

ALLOW/HOLD with machine-readable reasons (e.g., CT_INVALID, ACK_EXPIRED, CAP_APPLIED).

Change management

Save → transcript → canary → promote; replays must equal digest or no release.

Missing/expired attestations create deterministic HOLD rows with owners and SLAs.

Pricing & licensing that aligns to outcomes

Leg 1 — Deterministic Settlement (VRC)

Evidence-first feesTranscript & Digest parity

• Standard: 20% of Verified Recovered Cash (VRC) + $600 per payout window.
• Early self-funding: 80% of VRC to Verit until a $150–$200k cap, then 20% ongoing (+ per-window).
• All claims backed by Acceptance Matrix & audit-replay transcripts.

Leg 2 — Agentic AI (Automation)

Audited OPEX savingsMonth-12 license option

• Year-1 realized savings share: Client 80% / Verit 20%.
• At Month-12: optionally convert to a usage/license model (no revenue share).
• Metered by work performed with caps/bands; monthly utilization reporting.

Leg 3 — ML Optimizer (Spend Intelligence)

Invoice-proven savingsQuarterly reconciliation

• Realized savings share: Client 80% / Verit 20%.
• Savings attribution from provider invoices & benchmark deltas; optional floors/caps and category carve-outs.
• Designed to be net-positive within ~90 days of deployment.

Value PledgeEvidence-first • Customer-aligned • No ratchets

You only pay for what you save — and we will never increase our share, while you get more every day when we add more services to VeritOS.

Verified outcomes only (VRC • OPEX • ML invoices)Shares never ratchet upUpgrades included forever

See the ROI math Sales Partners exclusive access

Security, audit, and compliance — board-safe by design

Evidence without exposure

Content-addressed transcripts and digest equality let auditors verify correctness without raw PII.

Mapped to frameworks

Clear alignment to SOX 404, SOC 2, ISO 27001, PCI-DSS, GDPR Art. 5 with transcripted change control.

Operational recovery

Reason-coded holds, owner trails, and deterministic replays reduce incident blast radius and MTTR.

Ready to move from hope to proof?

We’ll run a discovery + replay pilot and deliver a proof-bound ROI plan tied to your numbers. Most programs reach go-live in about 90 days.

Play with an interactive Sandbox Watch a 3-min demo