Diarmuid Early

I’ve spoken to over 2 dozen MDs at PE firms I can confidently say that the arb of figuring out how to implement Vertical AI at portfolio companies is very real right now It will fundamentally change underwriting for those who can do it predictably and unlock generational returns. Most are aware they need to act. Very few have.

453

64 Comments

"The bad company problem likely originates in VC math. Too many founders are trained to look at market size or profit margin first. They see an opportunity that is large, or a company that could be highly profitable, and think “good enough.” This is because the early stages of any endeavor are risky. There has to be a large return to invest in something that early. It should create a 100x exit scenario. But that math isn’t the same as founder math. Founders, in my experience, tend to mistake “investing in” for “worth doing.” Being outcome-oriented is a mental disease. A plague. A pestilence. It strips the minute, daily beauties from around us and forces us to postpone happiness for “some day.” “I’ll enjoy the company once I fix growth.” “I’ll enjoy work once I’m post-money.” Maybe, just maybe, this is deeply unhealthy. I think part of the reason so many founders burn out, go through divorce, and have health problems is that there is a soul-deep misalignment between their tasks and their desires. The answer is so simple—the way to enjoy work is to do work you enjoy. The way to like your company is to make one you desperately want to exist. I don’t know why it has to be more complicated than that." https://lnkd.in/gdShJhXs

11

2 Comments

Perhaps more than any other internal discipline, the Finance & Accounting function often requires the most improvement among newly acquired SMBs. The company's new reality - which now includes a lender, an audit, and a Board of Directors, among other things - almost always requires new people, tools, and processes. This is often overwhelming for new CEOs, who are likely already consumed with countless other internal priorities. To help us make sense of what to prioritize, who to hire, and what to measure, I'm joined this week by Nicholas Andrews, CPA, Founder of the Aspen Consulting Group, a company that performs finance, accounting, and operations consulting for a wide range of SMBs. In our conversation today, we cover: - How much cash to keep on your balance sheet at any given time - Low friction ways to shore up liquidity, should it ever become a problem - Less obvious strains on cash flow that are often missed by CEOs - How much financial information CEOs should be sharing with their employees - What to do when your carefully crafted budget goes out the window in February or March - Whether you need a Controller or a CFO - How to make your first senior Finance & Accounting hire - ...and much more Thanks to Oberle Risk Strategies and Boulay for making this episode possible. https://lnkd.in/gPV9Jefw

22

Proposed QSBS update being considered for private company investors: “Taxpayers would be able to invest at later stages in startups and still get QSBS, cash out earlier and still get a partial benefit, and skip taxes entirely on up to 50% more of their windfalls.” https://lnkd.in/ge9_Vu-B

20

LP Direct Investments: The HealthVC Playbook Directs can close a fund or quietly hollow it out. This new issue gives you the operating system to run directs without hurting the fund. You’ll get: • Same-terms policy and a 72-hour timeline • Allocation waterfall that protects ownership and pro rata • Copy-paste emails for invite, decline, and post-allocation • Reporting lines for your LP letter • Risk map with fixes • A two-page policy and audit-ready deal log

11

Operating companies permanently changes how you invest. When you’ve lived through payroll cycles, commodity volatility, grid constraints, and board pressure in down markets, you stop starting with upside. Operators underwrite downside first: what breaks margins, what disrupts cash flow, what fails under stress. Capital doesn’t solve those problems but execution does. Operators: what risks do you underwrite first that others miss?

16

1 Comment

Too much of Private Equity is about allocating and financial engineering. It’s time to reinsert the operator experience and mentality into the control position. Everyone’s looking at compute budgets. We’re looking at who keeps the system running: - Subsystem welders. - Power component manufacturers. - Thermal design shops like our portfolio company TGM Systems. The people no one sees—but whose work makes modern AI infrastructure possible. AI Infrastructure Partners invests in that layer—the foundation where uptime, efficiency, and risk are managed in microns, not margins. Where a voltage ripple, not a missed KPI, is the real point of failure. This is where scale breaks—or scales. And where supply chain reliability becomes national security policy. We’re not chasing logos. We’re building industrial-strength platforms that serve the hyperscalers, defense primes, and frontier model builders deploying trillions in AI infrastructure—and needing to scale with conviction. That’s not venture hype. That’s foundational execution. It’s control capital for control systems. #SupplyChainAlpha #InfraStack #PrivateEquity #AIInfrastructure #BuiltNotBought #UptimeEdge #IndustrialDurability Jason Frank Damon Pitler David Henry, EMBA

26

8 Comments

Every day we speak to founders who've structured their companies exactly the way good advisors told them to — multiple entities, holding companies, intercompany arrangements — and then get declined because a lender called it "messy." Last week: a $5M SaaS business, 65% margins, zero debt, three entities, clean books, growing fast. Two lenders declined them — not because the business was risky, but because their systems couldn't untangle the structure. That's not a credit problem. That's a data problem the lender couldn't be arsed solving. Lenders confuse complexity with risk. And it comes down to tooling. Deterministic systems — rules, scorecards, thresholds — are great at consistency. Same logic, every time, fully explainable. Regulators love them. But they're brittle. Novel business structures break them. So the model says "decline" when the right answer was "dig deeper." Non-deterministic systems — ML, pattern recognition, probabilistic models — find signals humans miss. Margin compression months before a payment issue. Revenue seasonality buried in years of company-specific data. But "the model says no" isn't a credit decision. It's an abdication. You need both. At Tractor Ventures we've assessed hundreds of millions in loan applications to the borrowers most lenders call "too hard." Here's how: A deterministic scorecard analyses dozens of metrics across years of real-time company-specific financial data. Consistent. Auditable. Explainable. AI layers on top: normalising multi-entity structures, reclassifying messy financial data, flagging covenant risks before they become payment problems, generating the specific questions our credit team should be asking. Then a human makes the call — armed with better data and sharper questions than they'd ever prepare manually. Automate the quant. Amplify the qual. Keep humans where judgment matters. The result? A loss rate that would make most banks envious. And customers who come back over and over. Complexity isn't risk. It's a data problem we've solved.

110

8 Comments