The Deal That Closed Without a Deal Team
On July 1, 2025, The Pennant Group (NASDAQ: PNTG) announced it had acquired GrandCare Health Services, a home health provider serving Los Angeles, Orange, Riverside, and San Diego counties. Pennant's press release credited Montauk AI as placement agent. That single line did more work than most people noticed. A publicly traded acquirer, subject to SEC scrutiny and institutional diligence standards, closed a deal sourced and run by an AI-native investment bank with a headcount you could fit in a wardroom.
I ran an engine room on a submarine before I ran capital. Engine rooms do not care about your credentials. They care whether the system holds under load.
Montauk AI put its system under load against a public-company buyer and the system held. That is the story. Not the AI. The system.
Owner-operators in every industry, not just home-based care, need to study this deal. Your business is an asset. At some point you will need to convert that asset into cash, equity, or both. The mechanics of that conversion are changing faster than most advisors are willing to admit.
What Actually Happened
Strip the marketing language and here is the sequence. GrandCare, a Pasadena-based home health operator with three agencies and roughly 10,000 patients a year, engaged Montauk AI to run its sale process. Montauk AI was founded by Jarrett Bauer, a career home-based care operator who built and ran Health Recovery Solutions, a home telehealth company, for a decade before founding Montauk AI in 2023.
He is not a software founder who wandered into home health. He is a home health operator who built software.
That matters because the entire GrandCare transaction ran through Montauk AI's proprietary platform. According to the company's own account and independent coverage, buyers were given data portals that functioned as the investment memo. AI agents answered financial and diligence questions directly. Price alignment happened over software.
The deal closed well ahead of a traditional timeline. Pennant's Vice President of Acquisitions told Montauk AI the fit was strategic and cultural. The transaction closed with zero broad layoffs and an engaged staff post-close.
This is not a simulated case study or a hypothetical AI use case. It is a completed, publicly disclosed acquisition by a NASDAQ-listed strategic buyer, verified across the company's own materials, Pennant's investor relations page, GlobeNewswire, Nasdaq, StockTitan, and MarketScreener. The receipts are public. Anyone can pull them.
The Old Way: Six to Twelve Months and a Blank Check
Traditional sell-side M&A for a lower-middle-market company follows a doctrine that has not changed much in thirty years. You hire a boutique investment bank. You pay a retainer that runs $50,000 to $150,000 for full-scope engagement, sometimes with monthly work fees of $5,000 to $25,000 stacked on top.
You wait while a team of analysts builds a confidential information memorandum by hand, normalizes your EBITDA in a spreadsheet, and assembles a data room one PDF at a time. Then you wait again while they run outreach to a buyer list built mostly from personal relationships.
The whole process typically takes six to twelve months from engagement to close. Success fees on top of the retainer commonly land in the low-to-high single digits of enterprise value, with minimums frequently falling between $200,000 and $500,000 regardless of deal size. That fee structure exists because the workload is fixed.
Somebody has to manually build every diligence exhibit, chase every buyer question, and manage every counterparty by hand. It is not a scam. It is a system built around human bottlenecks, and it prices accordingly.
Compare that to what Bauer told Home Health Care News about a different Montauk AI client, Independence Home Health: a valuation and evaluation process that took two and a half weeks from start to finish. Not the full transaction. The evaluation phase alone, the phase that traditionally eats months on its own.
That is the compression owner-operators need to understand. It is not marginal. It is structural.
The AI Did Not Replace Judgment. It Replaced the Data Room Team.
Here is where I want to correct the record before someone oversells this. Montauk AI is explicit in its own research paper that humans still run the parts of the deal that require judgment, relationships, negotiation, and accountability. Analysts validate every AI valuation before a client sees it. Every client-facing call and negotiation is managed by a banker, not a bot.
What the AI agents replaced is the twenty-person back office that traditional banks staff for data room assembly, EBITDA normalization, and diligence Q&A logistics. Financial data gets ingested and normalized in hours instead of weeks. Buyer-seller fit gets scored in parallel across a curated database instead of sequential phone calls. EBITDA add-backs get surfaced systematically months before a process even begins, instead of getting discovered, or missed, during a frantic diligence sprint.
This is the same lesson I learned running a reactor plant. You do not want a human manually monitoring every gauge during a casualty drill. You want the system flagging anomalies instantly so the human can make the call that actually requires a human.
Montauk AI built the equivalent for deal data. The judgment stayed human. The grunt work got automated. That is the correct division of labor, one most owner-operators have never had access to because it required capital only Wall Street could afford to deploy.
The Owner's Exit Engine: A Framework You Can Use Today
I want to give you something more useful than admiration for a case study. Here is the framework I pull from this deal, doctrine you can apply whether you run a home care agency, an HVAC company, or a regional trucking outfit.
Compartmentalize your financials before you need to. On a submarine, you seal compartments before the casualty happens, not during it. GrandCare's clinical excellence and 5-star quality ratings became core to its valuation story precisely because the data existed in a form buyers could trust fast. If your books live in three disconnected systems and a bookkeeper's memory, you are not acquirable. You are a liability wearing a business's clothes.
Treat EBITDA normalization as a running system, not an event. Traditional owners normalize EBITDA once, right before a sale, under deadline pressure, which invites mistakes and buyer skepticism. Montauk AI's model surfaces add-backs continuously, months ahead of any process.
Build that habit into your monthly close now. Clean numbers are compounding the whole time you own the business, not just at the exit. By the time you are ready to sell, the story is already written and verified.
Understand that speed is now a competitive asset, not a luxury. Buyers penalize sellers who cannot produce clean data fast, because slow diligence signals operational risk. A platform that compresses evaluation from months to weeks is not just convenient. It changes who can credibly compete for the best buyers, because the best buyers move fast and walk from slow sellers.
Know the difference between an acquirable business and a sellable one. An acquirable business has clean financials, a defensible market position, and no single point of failure. A sellable business is acquirable and also has a buyer universe that can actually see it. Montauk AI's buyer-matching engine exists to solve that second problem for owners who never had Wall Street relationships. That gap has always been the real reason small operators got worse multiples than they deserved.
Build to sell from day one, even if you never plan to. A build-to-sell operator keeps the balance sheet clean, the systems documented, and the founder replaceable, because that is also how you run a stronger business while you own it. Sovereignty over your own numbers, not dependence on a broker's spin, is what actually protects your position in a negotiation. Owner-operators who wait until the exit conversation starts to get organized are negotiating from the back foot.
Your industry does not need to be home health for this to apply. The structural problem Montauk AI is solving, fragmented ownership and thin institutional support paired with buyers who reward data readiness, exists in home services, logistics, dental, veterinary, and manufacturing. The tools will follow the capital. Watch this model. It is coming to your industry whether you are ready or not.
Why This Is Not a Signal to Skip Due Diligence
I want to be direct about something the AI M&A hype cycle keeps getting wrong. Speed is not a substitute for diligence. It is a multiplier on it.
Montauk AI's own materials are clear that every externally delivered document passes human review and every negotiation runs through a banker. The AI agents did not eliminate diligence. They eliminated the manual labor that used to make diligence slow, expensive, and error-prone.
Due diligence is non-negotiable, whether it takes twelve months or twelve weeks. What changes is who does the grinding and how fast the grinding gets done. An owner-operator who reads this case study and concludes that diligence is optional now has misread the entire lesson.
The GrandCare deal cleared a NASDAQ-listed buyer's diligence bar. That bar did not move. The engine that met it got faster.
Montauk AI now reports more than $100 million in assets under mandate and is opening a Market Assessment to roughly 48,000 home health, home care, and hospice owners through September 29, 2026, with a stated mission to cut transaction times by 70 percent. That is one firm, in one vertical, out of an industry that includes roughly 30,000 home care agencies, 11,000 home health agencies, and 7,000 hospice providers nationwide.
The doctrine behind it is simple: compress the mechanical work, keep the judgment human, price the deal on verified numbers instead of hope. That doctrine is portable to any owner-operator sitting on an asset they have not yet learned how to sell.
Doctrine Connection: Due Diligence Is Non-Negotiable
The GrandCare-Pennant transaction is proof that automation and rigor are not opposites. Montauk AI's agents accelerated data assembly and buyer matching, but every valuation was validated by an analyst and every negotiation was run by a person accountable for the outcome. That is the doctrine.
Speed without verification is just a faster way to lose money. Speed built on top of verified numbers is a system worth trusting your balance sheet to. Build your own business so the numbers are already true before anyone asks.
FAQ
Q: Did Montauk AI really sell a company entirely through AI, with no human involvement? No. AI agents handled data ingestion, EBITDA normalization, buyer matching, and diligence Q&A logistics. Human bankers managed every negotiation and client-facing call, and analysts validated every AI-generated valuation before it reached a client or buyer, per Montauk AI's own published account of its process.
Q: How is this different from a regular business broker or investment bank? Traditional lower-middle-market investment banks charge a retainer plus a success fee, often in the low-to-high single digits of enterprise value, and typically take six to twelve months to run a full process because a human team assembles the data room and diligence materials manually. Montauk AI's platform compresses the mechanical portion of that work, which is why Bauer describes a two-and-a-half-week evaluation turnaround for at least one client.
Q: Can an owner-operator outside home health use this model? The specific platform is built for home care, home health, and hospice. The underlying doctrine, clean and continuously normalized financials, systematic buyer matching, and human-verified diligence, applies to any owner-operator business regardless of industry. Expect similar AI-native models to reach other fragmented, owner-operator-heavy sectors over the next several years.
Q: Does faster due diligence mean lower quality due diligence? No, and this is the point most people miss. Pennant is a publicly traded acquirer subject to institutional diligence standards. The deal met that bar. What got faster was the assembly and normalization of data, not the rigor applied to it.
Q: What should an owner-operator do today if they are three to five years from a possible exit? Start compartmentalizing your financials now. Normalize your EBITDA on a running monthly basis instead of scrambling before a sale. Build a clean, verifiable data trail so that whenever a buyer or a platform like this one comes calling, you are already acquirable instead of starting from zero.
Jeff Barnes has no personal position in any company named in this article. DEMG provides marketing systems, not investment advice.