Engineering Alpha in Private Equity
Aug 17, 2026 · 47 min · 9 segments
In Episode 13, Paul Karner and Dave Mangot sit down with tech veteran Michael Frendo, the current CTO of New Relic (a Francisco Partners and TPG company) and former executive at Proofpoint (a Thoma…
Michael FrendoGuest
Paul KarnerHost
Dave MangotHost
And then when I, you know, get to sort of my home turf, which is on the engineering side, it's really all come down to a, a, um, a total mindset shift in terms of how we develop technology versus how we have.
And, and, and to be fair, we're not there, we're not there yet, right? We're still, we're still evolving.
But then so is the technology, right? And so is the tech- You went back, uh, go-- to go back to a, a bit on a previous question around how do I get people on board, right? It's by painting.
Because where we are today with AI, I honestly believe we are just at the beginning.
We are just at the beginning, right? There is still work being done on how to, how to, how to build this capability so that it's more s- so that it's more distributed, so you can do more on the endpoint.
There's still work being done on how to get, you know, uh, a real reasoning model, not, you know, uh...
There's still work being done on how do you combine the, the probabilistic nature of how LLMs work with more, uh, deterministic skills that you can give it.
'Cause, you know, simple example, early days LLM, ask it to do math and it would fail, right? It, it couldn't do math.
If you said, "Do math using a calculator," it worked, right? S-simple, silly thing, right? But that's true i-in a broader sense, right? If, if I can do something, if I can provide the, the s- you know, the probabilistic engine with deterministic skills that it can figure out th- certain things.
I mean, yes, the model's getting bigger and it's getting smarter and they're training it, and it's learning from its mistakes, but it's al- they're also adding a lot of deterministic skills, right? Because if you can do something deterministically, you should, right?
I mean, it's almost 2010 in terms of, like, cloud, right? Like, what cloud looks like now-
it doesn't look anything like what cloud looked like in 2010, but that was three or four years into it.
I think the interesting thing I sort of picked up as a subtext from what you were saying was, um, you were talking about the people and getting them on board and bringing them along in the organization adop- But at no point did you talk about, "Well, now we're just gonna get rid of all the people, and we'll just have agents do everything," and then, like, "There's no room for people anymore here." And I, I don't know if that was an intentional subtext to what you were saying, but, like, it came through pretty strongly to me.
It's, it's-- Look, the list of capabilities and opportunities for us going forward is massive.
I mean, if anyone, like I said, uh, vi-vibe coding is not a path to building big and complex and successful systems, right? It is, uh...
And then when I, you know, get to sort of my home turf, which is on the engineering side, it's really all come down to a, a, um, a total mindset shift in terms of how we develop technology versus how we have.
And, and, and to be fair, we're not there, we're not there yet, right? We're still, we're still evolving.
But then so is the technology, right? And so is the tech- You went back, uh, go-- to go back to a, a bit on a previous question around how do I get people on board, right? It's by painting.
Because where we are today with AI, I honestly believe we are just at the beginning.
We are just at the beginning, right? There is still work being done on how to, how to, how to build this capability so that it's more s- so that it's more distributed, so you can do more on the endpoint.
There's still work being done on how to get, you know, uh, a real reasoning model, not, you know, uh...
There's still work being done on how do you combine the, the probabilistic nature of how LLMs work with more, uh, deterministic skills that you can give it.
'Cause, you know, simple example, early days LLM, ask it to do math and it would fail, right? It, it couldn't do math.
If you said, "Do math using a calculator," it worked, right? S-simple, silly thing, right? But that's true i-in a broader sense, right? If, if I can do something, if I can provide the, the s- you know, the probabilistic engine with deterministic skills that it can figure out th- certain things.
I mean, yes, the model's getting bigger and it's getting smarter and they're training it, and it's learning from its mistakes, but it's al- they're also adding a lot of deterministic skills, right? Because if you can do something deterministically, you should, right?
I mean, it's almost 2010 in terms of, like, cloud, right? Like, what cloud looks like now-
it doesn't look anything like what cloud looked like in 2010, but that was three or four years into it.
I think the interesting thing I sort of picked up as a subtext from what you were saying was, um, you were talking about the people and getting them on board and bringing them along in the organization adop- But at no point did you talk about, "Well, now we're just gonna get rid of all the people, and we'll just have agents do everything," and then, like, "There's no room for people anymore here." And I, I don't know if that was an intentional subtext to what you were saying, but, like, it came through pretty strongly to me.
It's, it's-- Look, the list of capabilities and opportunities for us going forward is massive.
I mean, if anyone, like I said, uh, vi-vibe coding is not a path to building big and complex and successful systems, right? It is, uh...
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.