Three bad answers, and a fourth one.
You finish the quality of earnings. The buyer thanks you, and then asks the question you have heard on every engagement: “So can I actually run this thing?”
Right now that question has three bad answers.
- 1.You decline it, and the buyer goes looking — which introduces another advisor into your client relationship.
- 2.You answer it informally on the call, which means giving an opinion outside your scope, unbilled, and on your professional liability.
- 3.You refer it to a systems integrator, who arrives selling an implementation and is not really doing diligence at all.
There is a fourth answer: a defined module, priced and scoped, that you attach to the engagement and deliver. Same data room. Same timeline. Same client relationship, still yours.
Concretely, it does three things for a QoE practice: it extends engagement value without extending accounting scope, it keeps a technical advisor from entering your client relationship on their own terms, and it gives your team something to say when the buyer asks a systems question in the closing call.
The questions it answers
- ?Do these numbers come out of a system, or are they rebuilt by hand every month?
- ?Who rebuilds them — and are they leaving at close?
- ?Can the buyer produce a defensible monthly close on day one?
- ?What actually breaks when the owner walks out?
- ?What systems is the buyer inheriting, and which are already at end of life or out of contract?
- ?How much operational and systems cleanup is the buyer buying, roughly, and in what order?
- ?Which of these findings should change price, structure, TSA terms, or the first 100 days?
What Aptum AI does
Reads the operating systems, data flows, and human dependencies underneath the reported numbers, and writes up what the buyer is inheriting. The deliverable is a written report with a source-of-truth trace, an owner-dependency map, a systems and technical-debt inventory, a reporting-readiness assessment, a ranked and costed post-close work estimate, and a deal-relevant findings section. Five to ten business days off the same data room. The six sections in detail.
Aptum AI was founded by Nick Graham. Twelve years in the U.S. Army, ten as a Green Beret. BA Mathematics, MS Finance, Stanford Ignite. Former VP of Operations at two venture-backed cybersecurity startups, and Senior Data Scientist on the Enterprise AI team at a publicly traded enterprise SaaS company. Selected to the inaugural cohort of Palantir's American Tech Fellowship for Veterans, 2026. Currently building a data warehouse, automated pipelines, and pricing-engine software for a PE-backed operating company in home services.
Prior portfolio work. Subcontracted through a partner agency: built a portfolio-wide operational KPI capture system for a private-equity owner of defense-focused IT companies.
What Aptum AI does not do
- —Any part of the Quality of Earnings. No normalisation, no adjustment schedules, no working capital analysis, no opinion on earnings quality. That is your work and it stays your work.
- —Any accounting, audit, or attest work.
- —Legal or tax diligence.
- —Cybersecurity assessment or penetration testing.
- —Valuation or fairness opinions.
- —Approaching your client independently. Introductions come through you, and Aptum AI will not solicit your client for other work during or after the engagement unless you say otherwise in writing.
- —Any guarantee about deal outcome, valuation, or investment return.
Four ways to run it
Referral
You introduce; Aptum AI contracts directly with the buyer and delivers under its own name. Cleanest from a liability standpoint. You stay informed; you carry nothing.
Subcontract
Aptum AI delivers to you under your engagement letter. You scope it, you bill it, you present it. Aptum AI is invisible to the client unless you choose otherwise.
Co-branded
The module is delivered as a named section of your report, with Aptum AI credited as the technical author. Useful where the buyer wants to know who did the technical work but you want to own the engagement.
Overflow capacity
Your team already does some version of this and is capacity-constrained in a busy quarter. Aptum AI takes the systems section on named deals.
Economics
Two structures, and the numbers are set with you rather than published:
Wholesale. Aptum AI quotes you a fixed delivery fee per deal, scoped on entity count and systems complexity. You mark it up and bill the client at whatever your practice supports. Your margin is yours and Aptum AI does not need to know it.
Referral. Aptum AI contracts directly with the buyer and pays a referral fee on collected revenue, agreed in advance in writing and disclosed to the client where your professional standards require it.
No rates, percentages, or volume tiers are published here on purpose. They depend on deal complexity and on which delivery model you choose, and quoting them before understanding your engagement economics would be guessing. The scoping call ends with a number.
White-label, independence, and liability
Under the subcontract model the report is delivered in your template, in your voice, with no Aptum AI branding, and Aptum AI makes no independent claim on the work. Under the co-branded model Aptum AI appears as the named technical author of a defined section and nowhere else.
Three things get settled before the first deal, in writing:
Independence. Whether your professional standards permit a subcontracted technical section under your engagement letter. That is your call, not Aptum AI's, and the referral model exists for when the answer is no.
Liability. Whose insurance responds, and what limitation-of-liability language sits in the subcontract. Aptum AI's liability is capped at the fee.
Attribution. Whether either party may reference the engagement afterwards, and in what words. Aptum AI's default is that nothing is referenced publicly without your written approval.
The objections you're already thinking
•“We already do a systems walkthrough.”
Most QoE teams do, at the level of “what ERP are they on.” This module traces individual reported figures back to what produces them and names what breaks at close. If your team is already producing a ranked, costed post-close work estimate, you do not need this.
•“Our independence rules won't allow it.”
Possibly. That is exactly why the referral model exists — Aptum AI contracts with the buyer, you carry nothing, and no independence question arises.
•“Our clients won't pay for another report.”
Then run it as a scoped-down three-day read on the next deal where the buyer is an operator rather than a fund. If they will not pay for it there, they will not pay for it anywhere, and both of us should know that quickly.
•“What if you are at capacity?”
Aptum AI takes a limited number of deals per month and will say no rather than deliver late. You will get a yes or a no within one business day of naming a deal.
•“Will you go around us?”
No, and it is in writing. Non-solicitation of your introduced clients, in whatever term length you want.
Practical details
•Does this touch our QoE workpapers?
No. Separate evidence, separate deliverable, no reliance either direction.
•Who talks to the client?
You, unless you decide otherwise per deal.
•How fast can you start?
Same week, subject to capacity. Name a deal and you get an answer within one business day.
•What if the buyer wants the post-close build afterwards?
Aptum AI will disclose that to you before quoting anything, and will not pursue it if you would rather it did not.
•Can we see a sample?
Yes — a redacted excerpt showing structure and finding format, on request.