For the CFO’s office
For the CFO’s office
Planning and budgeting. Contract management, physical and financial. And the exposure stated together — supplier, concentration, budget, and price. Read-only from the second week, through what finance already holds: the bill, the contracts, and the budgets.
What you get in two weeks
A first position, with its coverage stated. Every commitment you have given us with its floor, ramp, expiry, and drawdown pace. Every project budget against its spend. Every invoice we can match to its contract, with the exceptions first and the unmatched lines listed. The position: committed, forecast, and consumed, with what is above the floor and what it actually added to the bill. Prepayments outstanding, by supplier. Ninety percent of spend with a contract and rate card attached is the bar for the first print; automatic matching reaches its target by the third month. Nothing written to, and nothing that can sign, amend, or pay.
What you get with the job logs
The budget-at-risk: the probability of breaching the enterprise budget or any project’s, and what drives it. Next quarter’s commitment sized to measured demand rather than to a forecast nobody owns. And the desk’s daily record of what it would have bought next to what you paid — the number the subscription is judged against.
Planning and budgeting: the compute budget, held as a budget
| Function | What it does | Delivery target |
|---|---|---|
| Enterprise and project budgets | Authorized, reserved at approval, consumed as the work runs, remaining — by team and by project. The first notice is the request, not the invoice | Week 2 of the engagement |
| Commitment drawdown and sizing | The pace each floor is retiring at against what ran, and the shortfall or overflow it is heading for; next quarter’s commitment sized to measured demand | Week 2 of the engagement; sizing once job logs arrive |
| The forecast | What each piece of work costs this week and next, which decides where it runs; what next quarter costs, which decides the commitment; volume uncertainty included | Once job logs arrive |
| Budget-at-risk | The probability of breaching the enterprise budget or any project’s, from volume, the floors, overflow at on-demand rates, concentration, and price — and what drives it | Once job logs arrive |
| The marginal cash cost | What an hour actually adds to your bill under your contracts as they stand: the commitment it draws down or fails to, the credits and tiers your agreement applies to uncovered usage, the other services drawing on the same minimum, the term remaining. The rate card is an input, never the answer | Week 2 of the engagement |
| Accruals, prepaid drawdown, chargeback | Usage billed in arrears accrued; prepayments amortized as they draw down; credits and true-ups carried as their own lines rather than netted silently; cost allocated to the team and the purpose that used it | Week 2 of the engagement |
| Exports to the ledger and the treasury system | Billing exports in, in FOCUS format where your supplier publishes one; accruals, drawdown, and chargeback out to the general ledger as journal-ready lines, each with its lineage; the prepayment and commitment schedule and the exposure by supplier out to the treasury system. Your ERP stays the book of record; the record is a management record beside it, not a subledger | Week 2 of the engagement |
Built for a controller to close the month from, and for FP&A to forecast from drivers rather than guesswork. Every point of the bill committed and unused on a $60M fleet is $600,000, and every point of unplanned overflow is bought at on-demand rates — after whatever discounts your agreement extends to uncovered usage, which is still the worst price in the contract.
Contract management, physical and financial
| Object | What is in it | Delivery target |
|---|---|---|
| Contracts and commitments | Counterparty, floor, ramp, term, expiry, prepayment, credits, rate card, amendments, side letters — and the drawdown pace against what actually ran. Generation-swap and exit clauses tracked where they exist, and flagged where they do not | Week 2 of the engagement |
| Invoices, reconciled | Every invoice matched to its contract, rate card, and purchase order — in the contract’s currency, with credits, true-ups, and currency differences shown as their own lines — and, for purchases the desk ran, to the requirement written before the quote. Exceptions first, with the clause | Week 2 of the engagement |
| Price exposure | The position marked against all three index families, with the basis between them and your own contracts shown. We administer none of them | Week 2 of the engagement |
| Financial contracts | How much of the price exposure to fix, once a contract exists that can fix it and an accounting treatment exists to hold it. Always an overlay on a physical purchase, because a cash-settled contract cannot finish a training run | When a listed contract trades |
| The limits | The approval threshold, the concentration cap, the approved panel, the regions, and the budget the desk may draw on — the box finance draws around it | Week 2 of the engagement |
Counterparty credit risk, stated as a creditor would
CoreWeave reports prepayments averaging 15–25% of contract value across its active contracts at each of its last three year ends, and unless your contract secures it a prepayment is unsecured credit from you to a supplier whose own lease runs twelve years against a revenue book that is largely gone in four. The record qualifies every supplier separately on identity, technical capability, security, and financial resilience — from the filings where they exist: CoreWeave’s 30 June 2026 10-Q states total indebtedness of $35.6 billion against $5.5 billion of cash — shows prepayments outstanding, concentration, and the exposure cap against its use; screens the counterparty and the hardware’s location against the export-control lists, and tracks that law as it moves; and states a confidence level where a private supplier will not disclose, rather than printing a reassuring score. Whose loss is it if a supplier fails mid-run? Yours — and you are told before the placement, not after.
What it does not do
It prints no saving for holding the contracts — a record is not a saving. It carries no value-at-risk headline; it states the probability of breaching the budget and what drives it. It offers no financial contract until a listed one trades and an accounting treatment exists, and when one does, the desk prices it and finance decides. And it does not replace your ledger, your treasury system, or your procurement intake — it exports to them and ingests from them.
The demo for this seat
The record for an invented buyer, Helix Therapeutics: the position, the commitments and their drawdown, the project budgets and the approvals queue, the invoices reconciled with the exceptions first, the budget-at-risk, supplier exposure and counterparty credit, price exposure marked against three indexes, and the month close. Behind an email code; ask and we will let you in.