If you’ve never lost, you’ve never lived.

If you work in Private Equity, let's connect!

More than half of middle market portfolio companies now have some active AI initiative running. Fewer of them can tell you, in specific numbers, what that initiative actually changed operationally. Adoption isn't the differentiator anymore. Measurement is.

Entry multiples came down. Leverage came down with them. Which means the return has to come from somewhere else now, not the deal structure, the actual operation.That's not a talking point. It's arithmetic. Less leverage means operational improvement carries more of the return than it did five years ago.

Old model for scaling a portfolio company: fix the roadmap, then figure out demand. Better model right now: find out exactly where the operation already leaks time and money, then fix that first. The roadmap can wait. The leak can't. Measurement before strategy, not after.

Every portfolio company has two versions of itself. The one leadership describes in the board deck. And the one actually running on the floor, in the tickets, in the shift handoffs, in what the frontline team does when nobody's watching the dashboard. Most operational upside sits in the gap between those two. Not in a new strategy. In finding out which version is true.

Something I keep coming back to lately: "Play long-term games with long-term people." — Naval Ravikant

Jev ain’t for vibe coders

Random fact: in 2022 I was on my college's rocketry team, competed in a satellite competition in Virginia, placed 7th in the world. By the next year I wasn't building the rocket anymore, I was the liaison between the team and the organizers. Didn't realize it then, but that's basically the job I have now. Different rocket.

Open to anyone in the US, want to make some extra money on the side, no commitment attached? I run Drixal. We audit how a company actually operates, then implement the AI transformation that fixes what we find. Know a business owner who might benefit? Send them my way. If it turns into a paid audit, you get $500. No cap on how many. If they go on to a full transformation engagement, there's more on top. No interview, no application. DM me, I'll set you up with a simple way to track your referrals, and you're good to go.

Hot take: no generative AI in defense systems. Everything else is fair game. A classifier that's wrong fails in a predictable way, you can bound it, test it, know its error rate. A generative model that's wrong fails by sounding confident while being wrong. That's the failure mode you can't afford when the output decides who gets targeted.

If you run a business right now, which is actually true? -We know exactly where time and money leak =Rough idea, nothing measured =Honestly no clue =Never thought to check

Pressure is a privilege. No dependents, no mortgage, no one depending on my next paycheck. If I fail right now, I just fail. That's not a small thing.Most people don't get to risk it. I do, for now. So I am.

Is your company really ready for the AI era, or just using AI when it happens to come up? Made a free tool for business owners to actually find out. 14 questions, 5 min, gives you a score out of 100 Find out👇 score.drixalhq.com

Been a while since I did one of these properly. I’m Jay. Carnatic musician since 7. Studied CS almost by accident. Turned down a Deloitte offer after I already had it, to go build instead. First company, Wreak. Second, Steth, shut it down in October. Both had real product, neither had real PMF. What actually changed after that: I stopped planning a year out and started planning 3 months at a time. I stopped treating sales as someone else’s job and started doing it myself, badly at first, less badly now. I stopped building in silence and started building. Now building Drixal. Already stress-tested the system at real scale. Same person who builds fast. Different person about how he ships it.

Another day. Another model. Have you tried it yet?

A year ago, someone on an incubation panel told me: "You're really good with product. You don't know business, you don't know sales." Stung. I brushed it off. A year later, I know he was right on both counts. Still learning sales from scratch. No shortcuts.