Working with M&A advisors and accountants on AI evidence in DD
M&A advisors run the deal. Accountants verify the EBITDA. Neither is in the business of installing or verifying the operational AI stack that increasingly drives the multiple. That gap is where XLev sits - alongside the advisor, not against them - to make the AI story verifiable when the buyer's diligence team comes calling.
Three roles at the table
A sale process for a mid-market Australian SMB in 2026 has three professional roles working on the seller’s side, plus the operator:
- M&A advisor: positioning, marketing, buyer outreach, deal structuring, negotiation, transaction management
- Accountants and QofE provider: financial verification - the historical EBITDA is real, normalised and supportable
- Lawyers: contract, regulatory and IP review
- The operator: the business owner / CEO running the operation through the process
None of those four is the right party to install or verify a modern AI stack. That work sits in a fifth role - the operational AI partner - which is the role XLev plays inside sale-readiness engagements.
Why AI evidence matters in 2026
AI is increasingly part of the operating story buyers pay a premium for. Specifically:
- AI-driven productivity uplift shows up in the EBITDA the buyer applies a multiple to
- AI workflows that survive due diligence preserve that EBITDA; ones that do not get adjusted out
- AI governance and architecture maturity feed into the buyer’s view of risk
- The presence of a custom AI stack signals competitive moat versus an off-the-shelf operation
Buyers’ technical and operational diligence teams now ask AI-specific questions as a standard part of mid-market deals. The AI evidence pack is what answers those questions.
The AI evidence pack
The minimum viable AI evidence pack has three documents. Anything beyond is optional and adds value at the margin.
1. Workflow inventory
A structured list of every production AI workflow in the business, with for each one:
- Name and one-line description
- Function and owner
- Build cost (capex) and ongoing operating cost
- Wage-equivalent savings (hours saved per month, fully-loaded wage rate, net of platform cost)
- Vendor and platform dependencies
- Risk profile and any compliance considerations
2. Eval results and wage-equivalent savings
For each workflow, the operational evidence that the productivity claim is real:
- Baseline: hours per period before the workflow was AI-augmented
- Current state: hours per period after
- The wage rate used for the conversion (with source)
- Annualised savings, net of platform cost
- Eval results - the test set used to verify the AI's output quality, run on a regular cadence
A QofE provider can verify each of those inputs - which is exactly why the format matters. Without this evidence, AI productivity claims get adjusted out of QofE; with it, they survive.
3. Governance policy
The one-page AI policy plus:
- Training records showing every relevant staff member has completed the training
- Audit logs of the AI workflows (call logs, decision logs)
- Vendor contracts demonstrating the data-handling terms
- Breach reporting log (if any incidents have been logged - usually shows discipline rather than incidents)
How the handoff works
Two handoffs matter in a sale process:
- To the M&A advisor: workflow inventory plus the high-level wage-equivalent savings summary. This becomes part of the seller’s information memorandum and the data room.
- To the QofE provider: the full eval results, savings calculations, audit logs and underlying source documents. The QofE provider verifies the savings as part of the EBITDA quality work.
Both handoffs are easier when the AI evidence pack was assembled at the time the work was done, not retrofitted months before LOI.
When to bring an AI partner into the process
Ideally 18-36 months before sale. The AI evidence pack benefits enormously from being assembled in real time as the workflows are built and the operational evidence accumulates.
If you are already in advanced discussions with an advisor and a sale process is 6-12 months out, we can still help, but the value is highest when the operational work is already being done and we are documenting it as it happens. Trying to retrofit AI evidence in the last two months before LOI is hard and rarely supports the multiple as well.
We complement, not compete
We do not broker deals. We do not write QofE reports. We do not negotiate contracts. The M&A advisor, the accountant and the lawyer all do work XLev does not. Our role is operational - install the AI stack, document it, defend it under diligence - and that work makes everyone else’s easier.
For more on how we work alongside M&A advisors, see our piece on why XLev works alongside M&A advisors instead of competing with them.
For sale-readiness engagements specifically, the Sale-Readiness service page covers the Diagnose / Build / Stay phases in detail. Discovery call via the Contact page.
Frequently asked questions
- How does XLev work with M&A advisors?
- We sit alongside the advisor, not in competition with them. The advisor runs the sale process - positioning, marketing, buyer outreach, deal structuring, negotiation. We install and document the operational AI stack that drives a higher multiple, and we defend that work under buyer's due diligence. Most of our sale-readiness engagements involve close coordination with the seller's M&A advisor through the Build and Stay phases of the engagement.
- How does XLev work with the seller's accountants?
- We help the accountants assemble the AI portion of the Quality of Earnings narrative. AI-generated productivity savings need to be quantifiable, verifiable and supported by operational evidence - that is what the AI evidence pack is. We make the accountant's job easier by producing the artefacts they need in the format they need them. The accountant verifies the financial number; we verify the operational basis for it.
- Do you compete with M&A advisors?
- No. Different work, different skill sets, different parts of the deal. M&A advisors broker the deal; we make the business worth more before the deal happens. Most clients use both: us through Build and Stay phases of sale-readiness, an advisor through the transaction itself. The two complement; they do not compete.
- What is in the AI evidence pack?
- Three documents at minimum. Workflow inventory - every production AI workflow, what it does, who owns it, what it cost to build, what it costs to run. Eval results with wage-equivalent savings - what the AI workflow is achieving operationally, with baselines and current state, in $ saved per year. Governance policy - the one-page policy plus the training records and audit logs that demonstrate it is being followed. Optional extensions: the AI stack architecture diagram, vendor contracts, R&D Tax Incentive documentation if relevant.
- When should we bring XLev into the deal process?
- Ideally 18-36 months before sale. The AI evidence pack benefits enormously from being assembled at the time the work was done, not retrofitted in the months before LOI. If you are already in advanced discussions with an advisor about going to market, we can still help, but the value is highest when the operational work is already being done and we are documenting it as it happens.
Where this fits
Sale-Readiness
Specialised service. Operator-installed systems that make founder-led SMBs sale-ready in 6-36 months.