Sample brief. This is a real Deal OS run on a synthetic deal — Sentinel Fire & Safety, Inc. No real company or client data appears here. Every quote below was verified against the source document; claims that couldn't be verified were discarded before delivery (see the bottom of the page).

Project Sentinel — AI Diligence Brief delivered

Target: Sentinel Fire & Safety, Inc. — commercial fire & life-safety inspection, testing and maintenance (ITM).
TTM revenue $6.21M (+7.4% YoY). Adjusted EBITDA $1.42M (reported $1.08M).

Verdict: Attractive recurring base, but three findings materially change the risk picture and one stated fact is contradicted by the CIM's own customer detail. Proceed, with the items below confirmed before LOI.

Verified claims

Revenue growth is real and consistent — TTM revenue of $6.21M is up from $5.78M and $5.34M in the two prior years (~7.4% YoY), with no trend break. Confirm against monthly detail (not in the CIM): request trailing 24-month monthly revenue. high confidence
Source: Project_Sentinel_CIM.pdf · page 7 ✓ verified against source
“Total revenue for the trailing twelve months was $6.21M, up from $5.78M and $5.34M in the two prior years.”
Top-ten customer concentration is material and confirmed: the top ten accounts represent approximately 61% of revenue. This compounds the single-account concentration flagged below. Request 3-year revenue-by-customer to assess churn. high confidence
Source: Project_Sentinel_CIM.pdf · page 9 ✓ verified against source
“The top ten accounts together represent approximately 61% of revenue.”

Contradictions flagged

Two passages in the same document that cannot both be true. The brief shows you both — verbatim — and leaves the judgment to you.

critical contradiction The diversification claim is contradicted by the CIM's own customer detail
Page 8 states no customer exceeds 15% of revenue. Page 9's own figures put the top account (Northgate Property Group) at $1,366,000 against $6.21M total — that is 22.0%. The page 8 statement is false on the CIM's own numbers. Losing Northgate would remove roughly a fifth of revenue and a larger share of EBITDA. Treat concentration as a primary risk: verify Northgate's contract term, assignability on change of control, tenure, and relationship owner.
Project_Sentinel_CIM.pdf · page 8 ✓ verified
“Its customer base is well diversified, with no single customer representing more than 15% of revenue.”
vs
Project_Sentinel_CIM.pdf · page 9 ✓ verified
“Account A (Northgate Property Group) — $1,366,000; Account B — $544,000; Account C — $431,000; Account D — $388,000; Account E — $301,000.”
high contradiction The “78% recurring” figure is overstated by the Company's own definition
Page 7 presents 78% as agreement-backed recurring revenue. Page 22 reveals that one-time repair and call-out work is counted as “recurring” whenever it is performed for a contracted account. The two passages are inconsistent — true contractually-recurring revenue is lower, possibly materially. Request a revenue cut separating fixed inspection billings from variable repair/call-out billings. This affects the multiple a buyer should pay.
Project_Sentinel_CIM.pdf · page 7 ✓ verified
“Approximately 78% of revenue is recurring under multi-year inspection agreements, with the balance from repair, installation and emergency call-out work.”
vs
Project_Sentinel_CIM.pdf · page 22 ✓ verified
“Recurring revenue is defined by the Company as all revenue from accounts holding an active inspection agreement, inclusive of any repair or call-out work performed for those accounts during the period.”

Risks & items needing documentary support

critical risk Key-person risk is structural: the founder holds the license the business operates under
This is more serious than ordinary owner-dependency. If state law requires a qualified license holder to legally perform fire-protection work, the founder's retirement could leave the business unable to operate until a qualified replacement is licensed. Confirm the state requirement and whether any employee qualifies, or can qualify, to hold the license post-close. Potential deal-breaker that the CIM frames as routine.
Source: Project_Sentinel_CIM.pdf · page 14 ✓ verified against source
“The business is owned and operated by its founder, who holds the state-issued fire protection contractor license under which all Company work is performed.”
high missing info Contract assignability is unreviewed; concentration makes it urgent
Auto-renewing agreements are the thesis — but their value depends on surviving a change of control. With the top account at 22% of revenue, a non-assignment clause in the Northgate contract specifically would be major. The seller has not reviewed assignment provisions. Review the top-ten contracts' assignment clauses before LOI.
Source: Project_Sentinel_CIM.pdf · page 19 ✓ verified against source
“The seller has not undertaken a review of change-of-control or assignment provisions across the contract base.”
medium missing info Add-backs are itemized but two need documentary support
$340K of add-backs is 24% of adjusted EBITDA. Owner compensation and charitable items are defensible. The $65K “one-time legal settlement” must be documented as genuinely non-recurring, and the underlying matter understood. Request support for each line.
Source: Project_Sentinel_CIM.pdf · page 12 ✓ verified against source
“Reported EBITDA of $1.08M is adjusted to $1.42M after owner add-backs, including owner's compensation above market ($210,000), a one-time legal settlement ($65,000), personal vehicle and travel expenses ($41,000), and discretionary charitable contributions ($24,000).”
medium missing info Core operating data may not be exportable
Inspection schedules and compliance records are the operational backbone. An untested export from a 2009 proprietary system is integration and continuity risk. Test data export during diligence, not after close.
Source: Project_Sentinel_CIM.pdf · page 16 ✓ verified against source
“the seller notes that data export from this system has not been tested.”

What to confirm before LOI

  1. Revenue cut: fixed inspection vs variable repair/call-out billings.
  2. Northgate contract: term, assignability on change of control, tenure, owner.
  3. State licensing requirement + a qualified in-house successor.
  4. Documentary support for each add-back, especially the $65K legal settlement.
  5. Assignment clauses across the top-ten contracts.
  6. Tested data export from the legacy operating system.
  7. 3-year revenue-by-customer and staff turnover.
  8. 10-year litigation history.

Verification log — discarded before delivery

The model generated these while drafting. Each was checked against the source, could not be tied to a verbatim passage, and was removed before you saw the brief. This is the part most AI diligence tools never show you.

“The Company has no outstanding litigation.”
No passage states this. The page 12 settlement implies prior legal activity but supports no clean-litigation claim.
Surfaced instead: request 10-year litigation history.
“Average contract value is approximately $18,000.”
Cannot be derived from any single sourced passage; would require combining figures across pages the documents don't explicitly support. Unverifiable as stated.
“Technician retention is strong.”
The CIM lists team size (p.14) and certification (p.16) but says nothing about retention or turnover. No source.
Surfaced instead: request 3-year staff turnover.

Want this run on your own deal?

Upload a CIM and get a source-cited diligence brief — every claim verified against your documents, the unverifiable ones discarded.

Start a workspace →