Engineering Alpha in Private Equity

Cory Eaves on BayPine's operational playbook: targeting the analog-to-digital 'Goldilocks zone' and deploying frontline AI.

44 min Listen to the original Episode page
Private EquityArtificial IntelligenceOperational ImprovementChange ManagementDigital TransformationFrontline AIPortfolio Operations

Summary

Cory Eaves, Partner at BayPine, discusses the shift in private equity from financial engineering to operational value creation driven by technology and AI. He outlines BayPine's 'Goldilocks Zone' strategy, targeting non-tech companies with underutilized data and basic infrastructure. Eaves emphasizes bottom-up AI adoption via frontline tools, citing a partnership with OpenAI that improved customer experience and employment retention. The discussion also covers the challenges of change management, the resurgence of test-driven development for AI validation, and the long-term timeline for widespread AI transformation.

Key takeaways

  1. 1
    Private equity is shifting from financial engineering to operational improvement due to the end of the 'rising tide' market environment.

    Unidentified speaker (speaker_2) · 4:27

  2. 2
    BayPine targets the 'Goldilocks Zone': companies with existing data and core systems that are underutilized.

    Unidentified speaker (speaker_2) · 7:30

Episode sections

  1. 0:10 – 4:07

    Introduction and Cory Eaves Background

    Dave Mango introduces Cory Eaves, Partner and Head of Portfolio Operations at BayPine. Eaves outlines his background as a software CTO for three portfolio companies backed by General Atlantic before joining BayPine, highlighting his transition from tech operations to private equity.

  2. 4:07 – 7:13

    The Shift to Operational Value Creation in PE

    Eaves explains that the end of the 'rising tide' environment has forced PE firms to rely on operational improvements rather than just financial engineering. He notes that BayPine was purpose-built to partner with companies to use technology, data, and AI to change business trajectories.

  3. 7:13 – 9:47

    BayPine's 'Goldilocks Zone' Target Criteria

    Eaves describes BayPine's target profile as a 'Goldilocks Zone': companies with basic systems and organized data that are underutilizing them, avoiding both highly advanced digital companies and those with no infrastructure. They also seek management teams aligned with digital transformation and assess AI upside versus risk.

  4. 9:47 – 13:14

    Change Management and Founder Mindsets

    Eaves discusses managing resistance to change, noting that while founders are often open to digital transformation, technology leaders may be more resistant to abandoning legacy tools. He categorizes employees into three groups: enthusiastic, skeptical but open, and resistant.

  5. 13:14 – 20:32

    Analog-to-Digital Opportunities in Non-Tech Sectors

    The speakers discuss the vast opportunity in the 85% of the global economy outside of tech, citing examples like clinical trials running on paper binders and online education relying on PDFs. Eaves argues that middle-market companies lack the resources to modernize, creating a value-add opportunity for PE.

  6. 20:32 – 24:56

    Top-Down vs. Bottom-Up AI Adoption Strategies

    Eaves outlines two AI approaches: top-down process reengineering and bottom-up frontline adoption. He highlights BayPine's partnership with OpenAI to deploy tools to frontline employees, which led to unexpected innovations like using AI to diagnose and fix milling machine errors.

  7. 24:56 – 29:25

    AI Impact on Employment and Customer Experience

    Eaves argues that AI deployment at BayPine portfolio companies has been net positive for employment, with staff redeployed from administrative tasks to higher-value customer retention roles. He contrasts this with the 'replace everyone' narrative, emphasizing value addition through automation.

  8. 29:25 – 31:50

    AI Agents in Software Development and Security

    Speakers discuss using AI agents to review code for security and consistency, noting that while imperfect, these tools are superior to manual processes. They highlight the shift toward validating AI-generated code through automated testing.

  9. 31:50 – 33:24

    Validation Costs and Token Spend

    Dave Mango cites a claim that developers should spend twice as many tokens on validation as on code generation. He references a Concordia University study suggesting the vast majority of token spend is dedicated to validation.

  10. 33:24 – 39:29

    Citizen Developers and Prioritization Challenges

    The discussion shifts to 'citizen developers' and the difficulty of prioritizing projects when anyone can write code. Cory Eaves notes that some portfolio companies use AI-generated mockups to clarify requirements rather than pushing unvetted code to production.

  11. 39:29 – 41:58

    Change Management and User Adoption

    Eaves argues that technology implementation is easier than change management, citing a study where one-third of users admitted to sabotaging AI efforts. He emphasizes that aligning tech with business priorities and ensuring user adoption is the harder, more critical task.

  12. 41:58 – 44:04

    Long-Term AI Adoption Timeline

    The speakers agree that while AI is a once-in-a-lifetime opportunity, widespread company adoption and transformation will take years. They reference the Bill Gates/Benioff heuristic that people overestimate AI's impact in one year but underestimate it over a decade.

Claims worth verifying

Cory Eaves was a CTO for three different software businesses backed by General Atlantic.

Unidentified speaker (UNKNOWN) ·1:38

How to verify: Cory Eaves' LinkedIn profile and General Atlantic portfolio company announcements. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Primary source

Cory Eaves worked at General Atlantic for 15 years.

Unidentified speaker (UNKNOWN) ·2:27

How to verify: Cory Eaves' LinkedIn profile and General Atlantic press releases or biographies. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

BayPine was founded with the thesis of investing in non-tech businesses in financial services, healthcare, and industrials to modernize them with technology.

Unidentified speaker (speaker_2) ·5:54

How to verify: BayPine's website, pitch deck, or press releases regarding their investment strategy. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

The tech sector comprises approximately 15% of the global economy, leaving 85% as non-tech.

Unidentified speaker (speaker_2) ·17:45

How to verify: World Bank or IMF data on sectoral GDP contributions to the global economy. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

BayPine signed the first agreement between a private equity firm and OpenAI to roll out OpenAI to all portfolio companies.

Unidentified speaker (speaker_2) ·21:51

How to verify: Press releases from BayPine and OpenAI announcing the partnership. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Frontline employees at a BayPine manufacturing portfolio company used OpenAI to diagnose and fix milling machine errors, reducing downtime from days to minutes.

Unidentified speaker (speaker_2) ·22:47

How to verify: BayPine case studies, press releases, or internal reports from the specific manufacturing portfolio company. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

BayPine's AI deployment has resulted in net positive employment growth across its portfolio companies.

Unidentified speaker (speaker_2) ·26:22

How to verify: BayPine portfolio company employment data and operational reports. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

An online education portfolio company delivered courses via PDFs, contrasting with modern gamified expectations.

Unidentified speaker (speaker_2) ·15:34

How to verify: BayPine portfolio company website archives or case studies on their digital transformation. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Clinical trials for central nervous system drugs were conducted using paper binders and pen-and-paper notes.

Unidentified speaker (speaker_2) ·14:21

How to verify: BayPine case studies or operational reports from the specific clinical trials portfolio company. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Dave Mango gave a keynote at HGG Capital's CTO digital summit in Amsterdam.

Unidentified speaker (UNKNOWN) ·25:13

How to verify: HGG Capital event archives, Dave Mango's LinkedIn activity, or conference agendas. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Primary source

Dave Mango claims he heard someone say that if you're not spending twice as many tokens on validation as on writing code, you're doing it wrong.

Dave Mango ·31:50

How to verify: Search for the specific quote or source material cited by Dave Mango regarding token spend ratios in AI development. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Dave Mango references a Concordia University study stating that the vast majority of token spend is validation.

Dave Mango ·32:25

How to verify: Search Concordia University research publications or press releases for studies on AI token spend distribution. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Gene Kim, author of The Phoenix Project, discusses a renaissance of technologists returning to coding due to AI handling tedious tasks.

Dave Mango ·32:53

How to verify: Check Gene Kim's writings, blog posts, or interviews for discussions on AI's impact on developer sentiment and workflow. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

One of Cory Eaves' companies has embraced the 'citizen developer' model, using AI to create interactive mockups for business requirements.

Unidentified speaker (UNKNOWN) ·33:42

How to verify: Review BayPine portfolio company case studies or press releases detailing digital transformation initiatives involving citizen developers. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

A study cited by Cory Eaves found that one-third of users admitted to sabotaging AI efforts at their companies.

Unidentified speaker (UNKNOWN) ·39:50

How to verify: Identify the specific study referenced by Cory Eaves regarding user sabotage of AI initiatives and verify the statistic. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Cory Eaves states that AI adoption and transformation for companies will take a generation to play out, referencing the Bill Gates/Benioff heuristic.

Unidentified speaker (speaker_2) ·41:58

How to verify: Search for Cory Eaves' comments on AI adoption timelines and the Bill Gates/Benioff heuristic in BayPine communications or interviews. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify BayPine's partnership with OpenAI by checking press releases or public announcements from BayPine or OpenAI.
  • Review the Concordia University study cited regarding token spend allocation for validation versus generation.
  • Investigate the specific 'one-third of users sabotaging AI efforts' study cited by Cory Eaves to assess change management risks.
  • Check the financial filings of BayPine's portfolio companies to assess operational improvements post-AI deployment.
  • Research the definition and criteria of the 'Goldilocks Zone' target profile in BayPine's public materials.
  • Look for counter-arguments to the claim that AI deployment is net positive for employment in broader labor market studies.
  • Verify the claim that clinical trials for CNS drugs still rely heavily on paper-based processes through industry reports.
  • Assess the 'twice as many tokens on validation' rule of thumb against current industry benchmarks for AI development costs.
Full transcriptexpand
0:10
Unidentified speaker (speaker_0): From somewhere on the internet, it's Engineering Alpha and Private Equity. In this week's conversation, here's Paul Carner and Dave Mango.
0:20
Dave Mango: Thanks, Jay. Welcome back to another episode of Engineering Alpha and Private Equity. I am Dave Mango. and I am flying solo today. Paul is not able to make it for this recording. Um, but we're excited, both of us. I know Paul regrets not being able to be here, to welcome Corey Eves
0:40
Dave Mango: from Baypine to the program. I first bumped into Corey, met Corey, I don't know how you want to call it, when he was at General Atlantic, uh, and then he's been at Bay Pine for a few years. And I think one of the things I find pretty fascinating is the Bay Pine thesis
1:00
Dave Mango: is very highly aligned with the thesis of this podcast. Uh the idea that technology can really move the PL P L and get better returns and and things like that. And so um really happy to welcome Corey to the show. Uh welcome. And uh yeah, it'd be great if you could let the listeners know a
1:20
Dave Mango: little bit about yourself.
1:21
Unidentified speaker (speaker_2): Yeah, thanks, Dave. Happy to uh be with you guys or or you, I guess, uh today. Yeah, and dive in. Um
1:29
Unidentified speaker (UNKNOWN): thank
1:29
Unidentified speaker (speaker_2): yeah, thanks for having me on. Um
1:31
Unidentified speaker (UNKNOWN): yeah,
1:32
Unidentified speaker (speaker_2): let's see. I guess by way of introduction, uh I'm with Bay Pine, but maybe I'll start the other way around. Um
1:38
Unidentified speaker (UNKNOWN): start
1:38
Unidentified speaker (speaker_2): beginning. at the I'm a software uh guy by by background. I grew up in the software world um and uh eventually rose up and became a uh chief technology officer of a software business. Um I did that, my prior firm, General Atlantic Invested, did very well, sold it, recruited me to the next one, rinse and repeated. I did it
1:58
Unidentified speaker (speaker_2): again. Uh was there a few years, and then I did it actually a third time. they referred me to a third one. So I did sort of three tours of duty for them as a CTO. And then um they uh friend and a mentor of mine uh came over and sat down to dinner. And the first thing he said was, Hey, uh, I'm retiring. Do you want my job? Uh and
2:18
Unidentified speaker (speaker_2): so I, you know, overnight uh kind of switched from tech and that world into the private equity world. Uh
2:27
Unidentified speaker (UNKNOWN): I
2:27
Unidentified speaker (speaker_2): was back in oh eight, oh nine. I was probably the only guy in the world going into the finance uh, you know, industry at that point. And then um I joined General Atlantic when they were a kind of you know small growing firm. I was there 15 years and had an amazing, you know, career there at GA. And then uh three years ago I
2:47
Unidentified speaker (speaker_2): left um General Atlantic and was part of the early team here to help put Bape Pine uh together and I'm happy to share in a second here a little more about Bapeine. But that's the quick trajectory of, you know, kind of half half tech, half investing world. So
3:01
Dave Mango: I think the thing I love about that the most is, you know, there's uh the places where I'm reading most about private equity, I guess, would it be uh on LinkedIn, you know, other than newsletters from different, you know people who put on conferences or things like that. Uh, but a lot of the discussion on LinkedIn is
3:21
Dave Mango: about the need for people who have actually owned the PL or people who have actual experience inside of companies like being accountable to be like operating partners in private equity, because a lot of people maybe came from a consulting background where they jumped
3:41
Dave Mango: in, they jumped out, they made some suggestions, but they never actually were held accountable for the results. And what I like most about your story is being a CTO multiple times. You know what it's like inside what of these companies, you know what the people that you're talking to as an operating
4:01
Dave Mango: partner, you understand really. what they're going through. And to me, that feels like that would be a huge advantage.
4:07
Unidentified speaker (speaker_2): Yeah. Yeah. I I think so too. I mean it it's um private equity is going through like a whole really super important set of changes right now, which I can get into if if there's interest to to do it. But the net of it is that, you know, private equity firms have traditionally kind of been in this environment where you know the
4:27
Unidentified speaker (speaker_2): tide's rising, and when the tide's rising, every firm can do just fine. And now in a harder environment, uh harder investing environment, I think a lot of private equity firms are sort of realizing that they have to make operational improvements to their companies. You know, it's not it's not enough anymore to just buy something and hold on to it and sell it. You
4:47
Unidentified speaker (speaker_2): actually have to you know help the company grow and improve and change. And you know, that's what's really spurring um, you know, a lot of this hiring in in in private equity firms. Some firms get that better than others. You know, I I I I think there are some firms that uh really embraced it like ours and then other firms that that's maybe not quite so central to what
5:07
Unidentified speaker (speaker_2): they do. But nonetheless, that's the trend. I mean that's where the kind of whole private equity industry is headed.
5:12
Dave Mango: Yeah. I didn't I didn't let you know before we started recording. I have a soundboard and like I can make noises and like
5:22
Unidentified speaker (UNKNOWN): feel like
5:23
Dave Mango: for what you just said, like it definitely deserves some applause because uh
5:27
Unidentified speaker (UNKNOWN): it definitely is a crazy. I mean it's a good
5:30
Dave Mango: lead in to like BayPine, I guess.
5:33
Unidentified speaker (speaker_2): Yeah, sure. So I the BayPine Babine story in some ways um is kind of the embodiment of that. So
5:39
Unidentified speaker (speaker_2): I guess we've been around a few years, so I have to quit calling us a startup, but we're essentially a new private equity fund. And the font was founded with a really simple idea. As you kind of alluded to at the top, the idea is
5:54
Unidentified speaker (speaker_2): in essence, it's analog to digital. And maybe a little more specifically, it's is investing in non-tech businesses, really in three sectors. financial services, healthcare, and industrials, and then finding businesses that are kind of underutilizing technology. And then
6:11
Unidentified speaker (speaker_2): our firm is sort of built to partner with them, invest in them, and then use technology, data, AI to change the trajectory of the business. That's kind of what our firm was built to do. When we were getting the firm set up to do that, that was a completely radical idea. Uh, you know, other P firms were doing this a little bit around the edges, and we sort of said we're
6:31
Unidentified speaker (speaker_2): gonna like purpose build a firm to go to do this. Um, roll the clock forward to today. Uh, I think if you're an investor and you're not thinking about how technology is going to improve your businesses, you probably miss the boat. Uh,
6:44
Unidentified speaker (UNKNOWN): you know,
6:44
Unidentified speaker (speaker_2): it's people have kind of gotten the idea, but yeah, that's the the basic idea of of how a firm was was put together.
6:52
Unidentified speaker (UNKNOWN): So
6:52
Dave Mango: when you're looking at those targets for firms that you can improve, is there a specific profile that jumps out to Bay Pine that when you see, you know, these characteristics, you're oh, this is gonna be uh for uh you know a company that's actually well suited for the kind of work that
7:12
Dave Mango: we do?
7:13
Unidentified speaker (speaker_2): Yeah. Yeah. There's There a lot of things we'd look at. I might summarize it in in kind of three, you know, three broad buckets. One uh bucket for us is
7:22
Unidentified speaker (UNKNOWN): there's
7:22
Unidentified speaker (speaker_2): a little bit of a Goldilocks just right zone from a technology perspective. So when we see a company that um,
7:30
Unidentified speaker (UNKNOWN): you
7:30
Unidentified speaker (speaker_2): know, has got some of the basics in in place, they've got some core systems, they've got their data a little bit organized, um, but just, you know, it's not really fully utilizing it. That's actually a kind of a good good zone for us. There are a few companies we run into that they don't need our help. They're kind of already, you know, winning in the digital game, and that's probably not a great investment,
7:50
Unidentified speaker (speaker_2): you know, kind of profile for us. And there's some companies that are just starting at such a basic level where they have kind of absolutely nothing. I mean, it's hard to believe those companies exist today, but we do see some of those, and those are probably. not a great fit for us either. But if we can find a company that's got sort of good bones to it and a good basic, you know, um
8:10
Unidentified speaker (speaker_2): uh a good basic kind of foundation place, that's maybe one important criteria. Um the second thing we kind of look for is, you know, we look for in this AI world, one of the things we're always trying to judge now as it relates to AI is trying to judge
8:26
Unidentified speaker (UNKNOWN): kind
8:26
Unidentified speaker (speaker_2): of the AI impact on the business. And
8:29
Unidentified speaker (UNKNOWN): you
8:29
Unidentified speaker (speaker_2): know, everyone's very focused on AI risk. You know, is this business going to be disintermediated? Is this business going to have pricing pressure? And we absolutely give that, you know, a lot of thought. But the flip side is we also try and think a little bit about the AI upside. You know, is AI gonna unlock new markets? Is it gonna let this company launch new
8:48
Unidentified speaker (speaker_2): products? Um, you know, try and make a balanced view on how does AI affect the business. And then maybe the last one would be really management's vision for the company. You know, like we wanna partner with the company and invest behind a management team who sees this same digital perspective and, you know, opportunity
9:08
Unidentified speaker (speaker_2): for the business. We don't want to come in and be sort of convincing
9:12
Dave Mango: the
9:12
Unidentified speaker (speaker_2): management team that this is a good idea. We wanna want to like sort of preach to the converted a little bit and like find businesses where they see this opportunity in the business and they want to find a partner to um uh to help them go unlock it. So those those are some of the things we think about when we you know look at a look at a new business.
9:29
Dave Mango: Yeah I guess as someone who's uh who's gone in as the consultant on a lot of buyout sort of transactions uh that sometimes if the management team doesn't agree then then there's a new management team that does agree.
9:45
Unidentified speaker (speaker_2): Yeah.
9:47
Unidentified speaker (UNKNOWN): I've had
9:47
Dave Mango: enough uh engagements to their like, we'll call you back in a few months after things have settled down. I'm like, oh
9:53
Unidentified speaker (speaker_2): yeah. Yeah, I mean that happens. Um you know, our aren't it's not our preference. Ideally, you can find a management team that you're aligned to, you know, go go execute. on. But you know, there are some management teams or sometimes just um, you know, an individual or two who are
10:09
Unidentified speaker (speaker_2): resistant to change or don't, you know, have enough vision for what the future looks like. And in those cases, sometimes you you know do have to make a change. But um would
10:17
Dave Mango: found
10:18
Dave Mango: your like led companies be more likely to be like that because the founder sort of this is my baby, this is what I created. I don't want to change anything, I don't want to touch anything, it's still worth
10:29
Unidentified speaker (speaker_2): um I don't know. It's a good question. We do invest in a lot of founder businesses, and I would say, and I I think maybe this has to do with the fact that we're filtering for this, but I
10:38
Unidentified speaker (UNKNOWN): think
10:38
Unidentified speaker (speaker_2): a lot of our founders actually are kind of the opposite, you know, like a a lot of our founders are like a little their attitude is I've seen this uh potential sitting out there in in the business. and I don't quite know how to go execute on it. You know, like I I know we could do better and I see our
10:57
Unidentified speaker (speaker_2): competitors doing better. Um but I'm maybe not confident on how to go do it. I'm not too sure how to you know uh implement the technology. And most of our founders by and large um uh have been you know pretty embracing of it uh for the most part I it's a little more spotty. I think sometimes technology leaders
11:18
Unidentified speaker (speaker_2): um
11:19
Unidentified speaker (speaker_2): you know are the ones some not always but sometimes the ones that are a little more you know tied to yesterday's approach and you know sometimes take a little convincing or you know they have more invested in the way it was or something that you know those have probably been you know the years that Is we've had to work more on change management than than founders. Um so yeah
11:38
Dave Mango: I've I've done that with a lot of engineers. Or
11:42
Dave Mango: especially like moving from data center to cloud. They they don't want to get they don't want to give up their fancy tools that they convinced the company to buy because
11:52
Unidentified speaker (UNKNOWN): they got
11:52
Dave Mango: these fancy tools and now I'm an expert on this fancy tool and I'm like, that's great. It's just the whole industry has moved on from this.
12:00
Unidentified speaker (UNKNOWN): So
12:02
Unidentified speaker (speaker_2): Yeah, I I the way I sort of think about that a little bit is, you know, this three kind of buckets of people. There's usually you find some group leaders or engineers or, you know, business leaders who are they get the vision, they're enthusiastic about it, like they want to go execute on it. So that's easy.
12:21
Unidentified speaker (UNKNOWN): You
12:21
Unidentified speaker (speaker_2): get this next bucket, which are a set of people who are like a little skeptical but kind of be convinced. you know, like you could make the case, do proof of concepts, and kind of bring them along and kind of you know get them on board and so on.
12:33
Dave Mango: Somewhat curious.
12:35
Unidentified speaker (speaker_2): Yeah, somewhat curious and at least open-minded to, you know, be you know led in the right direction. And, you know, look, there's always this kind of third bucket of people that are just, you know, skeptics and just never gonna get there and never embrace the change. That's a little bit of just true agenda. in general. And um, you know, those the challenge there is just find those
12:55
Unidentified speaker (speaker_2): people early and help them find a better better role some somewhere else maybe uh you know, the the
13:02
Unidentified speaker (UNKNOWN): we're more
13:02
Dave Mango: suited to their approach.
13:04
Unidentified speaker (speaker_2): Yeah, yeah, exactly. Yeah. Yeah.
13:07
Unidentified speaker (UNKNOWN): Yeah.
13:07
Dave Mango: I think the interesting thing when you were talking about the profile of the different companies is
13:13
Unidentified speaker (UNKNOWN): uh
13:14
Dave Mango: I had a friend who worked uh for a while in like
13:19
Unidentified speaker (UNKNOWN): I
13:19
Dave Mango: don't know for lack of a better word, like digitizing shipping, not that they're shipping things digitally, but like the whole shipping industry
13:28
Unidentified speaker (UNKNOWN): and
13:28
Dave Mango: like how do you get things on boats and how do we maximize the you know the space utilization on a cargo ship and all all these kinds of things or whatever. And and he said something that stuck with me. He said, You'd be shocked how much of the world still runs on paper napkins that were signed at a bar in Singapore at
13:48
Dave Mango: two o'clock in the morning. And I that was a little frightening to learn. But uh are there companies that you bump into where you just kind of look at it and say, This is really a great opportunity.
14:03
Dave Mango: for digital, but also like the lift is just it's just too much. Like
14:10
Dave Mango: doing that.
14:10
Unidentified speaker (speaker_2): Um
14:11
Unidentified speaker (UNKNOWN): I
14:11
Unidentified speaker (speaker_2): don't think we've run into ones yet that that have you know kind of been in the too much category. We have a clinical trials business and this uh particular business
14:21
Unidentified speaker (speaker_2): um does executes clinical trials and they do it for a very specific class of uh central nervous system drugs. And I can remember they have a uh number of sites around the country and I went to tour a couple of the the sites with the management teams. And I can remember them sort of opening this room and then just seeing row after
14:41
Unidentified speaker (speaker_2): row after row of paper binders. You know, like they had a each patient sort of had a binder and then the nurses in the center were like taking notes with a pen the and paper. in the binder for the patient. And you know, part of you sort of looks at that, you know, it's a little incredulous, like, wow, you know, this is the healthcare system running on
15:01
Unidentified speaker (speaker_2): pen and paper. But you know, the other part is like, man, talking about low-hanging fruit. I mean, just like,
15:06
Unidentified speaker (UNKNOWN): you
15:06
Unidentified speaker (speaker_2): know, there's so much that could be done there that, you know, like that's what inspired us to really make make an investment in that sort of business. I mean, so that'd be one example. I mean, another example would be we own a um uh education business, online education business. And you know, when we
15:26
Unidentified speaker (speaker_2): first uh invested in that business, their kind of idea of online education was
15:34
Unidentified speaker (speaker_2): kind of like emailing you a PDF. You know, and if you think about what people expect today from an online education business, I mean you probably have Duolingo loaded on your phone or something, right? And it's
15:45
Unidentified speaker (UNKNOWN): like
15:45
Unidentified speaker (speaker_2): this,
15:46
Unidentified speaker (UNKNOWN): you
15:46
Unidentified speaker (speaker_2): know, like this highly interactive, gamified system. It's in little bite-sized pieces, it's super compelling. Like people today in their own lives have an impression of what that that should look like. And if you sign up for a course and they suddenly email you this hundred-page hundred page PDF and say call back in six weeks for an exam,
16:06
Unidentified speaker (UNKNOWN): you
16:06
Unidentified speaker (speaker_2): know, that's not a very compelling user experience
16:10
Unidentified speaker (UNKNOWN): in
16:11
Unidentified speaker (speaker_2): that,
16:12
Unidentified speaker (speaker_2): you know, case. And that's one where we can completely revamp that whole business to, you know, bring that to what you'd expect from a modern experience. So yeah, there's you just see examples like this to too, your, you know, napkin example at Ship Bang. I mean, it's uh it's kind kinda of everywhere you turn. y you see these examples of opportunity out there to to go do this.
16:31
Dave Mango: Speaking of opportunities, I missed when you said you probably have Duolinga on your phone, I should have said C Claudo or something like that, but I I
16:40
Unidentified speaker (UNKNOWN): missed my
16:40
Dave Mango: opportunity for the
16:42
Unidentified speaker (UNKNOWN): for the joke.
16:43
Dave Mango: It only hit me later. Um
16:45
Unidentified speaker (UNKNOWN): Yeah.
16:46
Dave Mango: I did a due diligence for um I guess it was a marketing company. and they were sending contracts back and forth to all these advertisers, but it was all like PDF and just it was basically just paperizing like you know, taking what had been paper and making it digital paper
17:06
Dave Mango: because it was PDFs. And the the way that the team like figured out to start to handle these things in bulk was they put QR codes on all the PDFs so that when they would get the signed PDF back, the computer could just read the QR code and all of a sudden it would know exactly which document that was
17:26
Dave Mango: and I was like, that is awesome. Yeah,
17:30
Unidentified speaker (speaker_2): that's a great stuff. Stuff. And look, I think
17:33
Unidentified speaker (speaker_2): I think this opportunity is well, put it this way, one of our founders talks about this occasionally. You know, if you take the tech world, the world that you and I are familiar with,
17:44
Unidentified speaker (UNKNOWN): hardware,
17:45
Unidentified speaker (speaker_2): software, services, all of it, you just take the whole tech universe and add it up, it's about fifteen percent of the global economy.
17:53
Unidentified speaker (speaker_2): The other eighty five percent of the global economy is everything else. You know, it's shipping, it's manufacturing, it's education, it's government. There's just this like this massive part of the economy. And I think for people who are kind of grown up in the tech world, you sometimes look at those other parts and it's a little like, wait, wait a second, how does
18:13
Unidentified speaker (speaker_2): this kind of like still work over here like this? And you know, that's to some degree, that's a little bit like the core reason PayPal pay fine exists, is there's just so many businesses. And I would also say, especially in uh um
18:27
Unidentified speaker (UNKNOWN): the
18:27
Unidentified speaker (speaker_2): middle market, you know, like if you're a big company in these things, you have the resources and talent, leadership to go tackle some of these things. But a lot of middle market companies just don't have access to those things. So um so yeah, no, I'm with you. I think there's there's a lot of opportunity to go go focus on this stuff for sure.
18:46
Dave Mango: Um so obviously we can't do a podcast without talking about AI because you know we that would be like some kind of
18:55
Unidentified speaker (UNKNOWN): I think doesn't
18:56
Unidentified speaker (speaker_2): every business meeting have to talk about AI now? Uh uh but probably. Yeah.
19:01
Dave Mango: Um but I guess like you know, Bay Pine has been around for from before I g you know AI became this huge thing. Certainly, you know, there were a there was ML, they were you know, those are those kinds of things that were around. But specifically in like these businesses that
19:22
Dave Mango: are, let's say, lagging. So like you kind of named like you know, government and all these other things, you know, clearly like, and I have some thoughts about some of this, but you know, So we also want to make like the experience better for like the people who are the customers. of these businesses, and you know that it's not
19:42
Dave Mango: just about like okay, now we can do stuff faster, like okay, so I can do it faster, who cares? Um, like what kinds of things are you seeing like are good uses of AI? And so I'm sort of thinking about this in the specific lens of I can't remember which PE firm it is, but like they're they're doing a lot of
20:02
Dave Mango: bringing AI. into traditionally services businesses. And what they do is they sit down with the people who are working in those businesses, in those services jobs, and really deeply try to understand the work that's going on there instead of just saying, Oh, yeah, here, we'll just rub some AI on it, everything'll Everything will be fine.
20:22
Dave Mango: And then, you know, I have a feeling you're gonna say, like in three seconds, everything's not fine. when you do that. And
20:28
Unidentified speaker (speaker_2): yeah,
20:28
Dave Mango: um, like how are you approaching like those kinds of situations?
20:32
Unidentified speaker (speaker_2): Yeah. Well,
20:34
Unidentified speaker (UNKNOWN): yeah,
20:34
Unidentified speaker (speaker_2): there's a lot to be said on this. I I I maybe start here, Dave, and say we there's sort of two ways to think about this, and you know, maybe maybe more like one one way is like a big tops down sort of project that you can run at these at these businesses, and we've done that. I'll give an example or two here in a second. But you know, these are projects
20:54
Unidentified speaker (speaker_2): where you get alignment at the top, you scope it out, you kind of figure out where AI can be used in the in the business. And those are um those are the things that can be real uh needle movers you know on on the business. Um they often entail uh a lot of process reengineering,
21:12
Dave Mango: you
21:12
Unidentified speaker (UNKNOWN): know,
21:12
Unidentified speaker (speaker_2): where it's like not just taking apart one little piece. of the puzzle, but like taking a more holistic view. Yeah.
21:17
Unidentified speaker (UNKNOWN): And it
21:18
Unidentified speaker (speaker_2): also takes a lot more change management, you know, like to your point of like sitting down with employees and helping them, you know, redesign their role and stuff. Um, so we're doing those. We're doing a bunch of those. We're doing those at almost every single business that that that we've got.
21:31
Unidentified speaker (UNKNOWN): The
21:31
Unidentified speaker (speaker_2): other approach, which is a little more the bottoms up approach, is valid too. Uh and you know, one of the things we did about a year ago, was we actually signed the very first agreement between a private equity firm and OpenAI and formed a partnership to roll OpenAI out at all
21:51
Unidentified speaker (speaker_2): of our portfolio companies. So we actually bought thousands and thousands of seats of OpenAI and really put it in the hands of the frontline employees out in the businesses. And you'd be amazed at some of the things that the employees just sort of found as use cases
22:11
Unidentified speaker (speaker_2): to, you know, on their own to do with those things. And now look, those things are not no one single one of those is a you know gonna move the needle. But if you take thousands them across
22:22
Unidentified speaker (UNKNOWN): you know a company,
22:22
Unidentified speaker (speaker_2): they they could um an example I love here is um we have many a manufacturing company. It has these large complicated milling machines uh in the business. And when one of these things historically would break, something would go wrong, it'd you know show up an error code or a diagnostic and the company would call
22:42
Unidentified speaker (speaker_2): the manufacturer and they would come and service the the the machine.
22:46
Dave Mango: Yeah.
22:47
Unidentified speaker (speaker_2): We put OpenAI in the hands of the uh of the employer the employees and the employees sort of figured out completely on their own that the on their phone they could take a screenshot of the machine and understand the error code and take a machine a shot of the milling thing and understand what was happening. And they themselves could
23:07
Unidentified speaker (speaker_2): actually fix the machine and get the machine back
23:10
Unidentified speaker (UNKNOWN): online.
23:11
Unidentified speaker (speaker_2): Right. And what would have taken three or four days to call somebody out to come up you know, fix this machine. Now these guys have sort of figured out on their own, like how to do this basically kind of on the spot. It's a really cool example because if you had thought about that, you probably would have never thought, oh, this is something we're gonna, you know, go do in this business. But
23:30
Unidentified speaker (speaker_2): the frontline workers just kind of figured it out on their own, you know, uh to go to go do that, which I we have tons of those examples. but I kind of love those, you know, those kinds of things. So
23:43
Unidentified speaker (UNKNOWN): the machine
23:43
Dave Mango: is down, that's just wasted m you know, time and money.
23:47
Unidentified speaker (speaker_2): Yeah, yeah, for sure. Yeah. So we're we're we're doing both of those, um, both of those things in in the portfolio. Look, I I I would say the um in in terms of your other part of your question, which was like where are we seeing lots of usage, this will be no surprise, but um the the right tip of the spear is software engineering,
24:07
Unidentified speaker (speaker_2): software development. You know, all of our companies are uh doing some level of custom software development. And and as you know, I mean the just the the games and the productivity and the speed of the tool change is just like mind blowing in that space. So that's that's a 100%. hundred percent. Like every company is doing software engineering is using.
24:27
Unidentified speaker (speaker_2): the tools in in that way.
24:30
Unidentified speaker (UNKNOWN): A
24:30
Unidentified speaker (speaker_2): lot of work in call centers, you know, and like contact centers and stuff has been a you know another real common area where, you know, it's been pretty, it's not even really that difficult anymore to like automate pieces of, you know, contact centers and agents, chat agents and that sort of sort certain of thing. things. So yeah, we're definitely seeing some of these areas that
24:50
Unidentified speaker (speaker_2): are reusable and, you know, we're doing over and over again in the in the portfolio.
24:56
Dave Mango: Yeah, I I I'm like I'm like in my head I have two different directions to go from there. But uh
25:03
Unidentified speaker (UNKNOWN): I
25:03
Dave Mango: gave I guess the the thing about the call center is I gave the keynote for H HG G Capital's CTO digital summit or whatever a few years
25:13
Unidentified speaker (UNKNOWN): ago
25:13
Dave Mango: in Amsterdam. And
25:15
Unidentified speaker (UNKNOWN): I
25:16
Dave Mango: was listening to one of the other, you know, presentations after me,
25:20
Unidentified speaker (UNKNOWN): and
25:21
Dave Mango: they were talking about call centers, and they said the root thing that they want to use AI for in the call centers isn't to get rid of all the call center agents. It was there's so many questions that come in that are just easy to answer, and they're wasting their experience. experienced agents' time on these
25:41
Dave Mango: easy questions. And they would much rather have an AI tackle those things so that their big customers who are spending a lot of money can get real personalized service that enables them to take advantage of the platform much better than they ever were before. And that was the way that they were thinking about it. And I
26:00
Dave Mango: was both shocked and impressed. Because I for me, uh like there's been so much about oh, we're just gonna replace everybody with AI, and that was a completely different like approach. It was yeah, we're going
26:15
Unidentified speaker (speaker_2): to make
26:16
Dave Mango: them team members and help instead of you know get rid of everybody and replace them.
26:22
Unidentified speaker (speaker_2): Yeah. I I I echo that. I mean if I would look if I look holistically across our companies, I would say at this point, um we are actually net positive with new employment in the company. So even if even as we've deployed all this AI, our companies are growing and scaling and launching new products and going into new markets and so on.
26:42
Unidentified speaker (speaker_2): And you know, they're adding they're adding people. And like to pick us a s another other specific example, this education business I mentioned,
26:49
Unidentified speaker (UNKNOWN): you
26:49
Unidentified speaker (speaker_2): know, they have a large contact center and they had people who were, you know, a student would enroll in the school. And one of the things you had to do was go get all their old transcripts and make
27:01
Unidentified speaker (UNKNOWN): sure that
27:01
Unidentified speaker (speaker_2): the credits transfer over. And
27:03
Unidentified speaker (UNKNOWN): that's a
27:03
Unidentified speaker (speaker_2): pretty
27:04
Unidentified speaker (UNKNOWN): tedious,
27:05
Unidentified speaker (speaker_2): you know, mechanical thing, probably brings back flashbacks for all of us of uh, you know, high school and college. Um, and we automated a ton of that. So like now that automation sort of happens. And those same people who were doing this kind of you know administrator type work are now actually counseling employ counseling students where
27:25
Unidentified speaker (speaker_2): they're calling up students and sort of saying, Hey, we noticed you're falling a little behind.
27:29
Unidentified speaker (UNKNOWN): Um,
27:30
Unidentified speaker (speaker_2): you know, how is it that we can help you get back on track? You know, they're they're doing work that's actually more valuable for them, like it's more interesting, you know, in engaging sort sort of work for them. It's actually better for the student, uh as well in that, you know, environment. So yeah, I mean I I think your example is like spot on. I mean that that's absolutely been the the experience that
27:50
Unidentified speaker (speaker_2): we've seen in the in the portfolio. Yeah.
27:53
Dave Mango: Yeah.
27:54
Dave Mango: Okay.
27:55
Dave Mango: I have a lot of thoughts. I uh
27:59
Dave Mango: I I just like this whole idea of uh the AI AI's
28:04
Dave Mango: making the company more valuable through addition instead of like replacement. And so uh the one I've come up with this week, which uh as a former CTO I think you will appreciate is uh
28:18
Unidentified speaker (UNKNOWN): back
28:18
Dave Mango: in the day everybody and all the uh
28:21
Unidentified speaker (UNKNOWN): I
28:21
Dave Mango: guess w we didn't have CISOs for a while. Now we have CISOs, but like there's whole idea in security of we're gonna have security champions and they're gonna be these developers who are on the Scrum teams and they're going to know slightly more about security than other people. And then the DevOps people
28:41
Dave Mango: were like, Oh, this is a great idea. We're gonna have DevOps champions, and that it's basically like outsourcing all of your job to like people on the developers instead of you having to do it. And you know, as you might expect, the um
28:55
Dave Mango: the quality, let's say, of this is gonna be extremely uneven. And everyone I've talked to pretty much who has had these champion programs eventually they folded, which I don't think is quite a big surprise. But um you were talking about open OpenAI, AI, and open OpenAI AI recently published how they're
29:15
Dave Mango: doing their sort of agentic development uh at OpenAI. And one of the things that they do is they have a security agent that reviews the code,
29:25
Unidentified speaker (UNKNOWN): and they
29:25
Dave Mango: have a cloud infrastructure agent that reviews the code. And I was like, oh my gosh, this is your champion idea, except it scales across the entire business, and it's
29:35
Unidentified speaker (UNKNOWN): something that
29:36
Dave Mango: we've always tried to do in the past, except we did it terribly. in the past, and we did it.
29:41
Unidentified speaker (UNKNOWN): And
29:41
Dave Mango: now you can have this,
29:42
Unidentified speaker (UNKNOWN): you
29:43
Dave Mango: know, even if they're not the best security, whatever in the world, like it's still way better than not having them. I mean, unless you really screw it up, I guess. But
29:52
Unidentified speaker (UNKNOWN): uh
29:53
Dave Mango: I really am super fascinated now with this idea of what are all the things that we always wanted to do in software development
30:01
Unidentified speaker (UNKNOWN): that
30:01
Dave Mango: we couldn't do between. because it was cost prohibitive or this or that, and now we can assign agents to some of these tasks and make everyone's job easier and better.
30:13
Unidentified speaker (speaker_2): Yeah. Yeah, I I'm love to think about this kind of stuff too. I'm the engineer in me loves loves brainstorming about that stuff. I um
30:22
Unidentified speaker (speaker_2): uh
30:23
Unidentified speaker (speaker_2): and yeah, uh
30:26
Unidentified speaker (UNKNOWN): talking
30:26
Unidentified speaker (speaker_2): about change. I mean the amount of change that AI even in the last few weeks has pushed on this security world is
30:33
Unidentified speaker (UNKNOWN): about
30:33
Dave Mango: to say you're talking about security vulnerability.
30:35
Unidentified speaker (speaker_2): Yeah, it's sort of crazy. That's a whole I'm sure podcast by itself. Um but yeah, I mean I'm I'm working on kind of a side project on the side of my desk here uh for an internal thing at Baitmind. Baebind. And uh I wrote 18 different agents that then review the code for security,
30:55
Unidentified speaker (speaker_2): for completeness, you know, for duplication, for architectural consistency. You know, and every time I do a check-in on the code I'm writing, uh, where AI's writing, I guess, um, you know, these agents scan the code and provide feedback on the code. And I think that's yeah, absolutely, you know, kind of part of how it works. And again, not to geek out
31:15
Unidentified speaker (speaker_2): too much, but you know, test driven development was a trend
31:19
Unidentified speaker (UNKNOWN): for
31:19
Unidentified speaker (speaker_2): a while, though, which I think we're sort of getting back to, uh dealing with software development. I mean, it's easy to create all this code with AI, but unless you have some way to like validate it, make sure that it's, you know, doing what you think it's going to do, So uh you know, it's just impossible for humans to keep up with that that stuff anymore. So yeah no, it's
31:39
Unidentified speaker (speaker_2): fascinating. I I there's just the the AI uh impact on software development, all all every piece of software development is just kind of mind blowing at at this point. Yeah.
31:50
Dave Mango: Yeah, I um I don't know if I agree with it a hundred percent, but I heard someone say last week that if you're not spending twice as many tokens on validation as you are on writing the code itself. You're doing it wrong.
32:04
Unidentified speaker (speaker_2): Oh, that's interesting.
32:05
Dave Mango: And
32:05
Unidentified speaker (UNKNOWN): yeah.
32:06
Dave Mango: It's all this idea of like how, you know, this is how agents work. They they keep writing the code until it the test pass, but like there's so much of this is validation. And that's where the real and I've seen other studies. I think Concordia University had one that basically back I don't think it was twice as many,
32:25
Dave Mango: but like it was
32:27
Unidentified speaker (UNKNOWN): it
32:27
Dave Mango: definitely said like the vast majority of token spend is validation. Um but also
32:33
Unidentified speaker (UNKNOWN): I
32:33
Dave Mango: mean the kinds of things that you're talking about, number one, uh I think I've not just me, but like uh Gene Kim, the guy who wrote the Phoenix Project, talks about this sort of renaissance now of all these technologists who have gotten away from code. and all this other stuff over their careers
32:53
Dave Mango: are now coming back and like re experiencing the joy that they had when they were
32:59
Dave Mango: younger doing all this stuff because now it's not tedious. Now it's the creation that they always loved, except like now the agents are doing a lot of this, and it's sort of bringing all these people back into the fold. And then, you know, you as like a former science. software and CTO and you're like, oh and I also have a side
33:19
Dave Mango: project I'm doing. I'm like, yeah, that tracks that tracks pretty good.
33:24
Unidentified speaker (speaker_2): Yeah. Yeah, I think I can identify with that. You know, um it's it's captures me. I mean I I'd also say um on a s similar, not exactly the same theme, but you know we have one of our companies that's really like super embraced this citizen developer um, you know, kind
33:42
Unidentified speaker (UNKNOWN): of
33:43
Unidentified speaker (speaker_2): idea. Um it's a little similar, not quite the same idea, but it you know, this idea of like take people who have a passion for the business and have just like a certain degree of technical capability, but maybe weren't coders. Um and you know like this other business of ours has really started to figure out a way to harness that, um, which has been like super interesting. to
34:03
Unidentified speaker (speaker_2): to see that work too. So
34:06
Dave Mango: good. This is a good one I want to get your perspective on because I I'm seeing this a lot in the portfolio companies that I talk to is this idea of I don't know what they call it, democratization of development. I guess election day is next week in the US and the Israelis and the Serbians and the Brazilians all just had
34:25
Dave Mango: elections day. So we can we can say democratization, but uh
34:30
Dave Mango: but part of the thing that I'm struggling with when I'm talking to these companies, and I I'd love to get your perspective on this, is if we do have all these citizen developers and everybody can write code, how do we decide what to build and how do we prioritize this stuff? Because you know, throwing garbage at the
34:51
Dave Mango: wall faster doesn't actually make the business more valuable. And like not
34:56
Unidentified speaker (UNKNOWN): all
34:56
Dave Mango: ideas are are well researched or good. And I I'm not trying to, you know, throw a wet blanket on innovation in any way, but also like, you know, Edison tested what, 10,000 thousand f filament things. I'm sure there were a lot of things where he was like, we're not testing that. because it's not even conductive. So
35:16
Dave Mango: like let's not bother. And
35:18
Unidentified speaker (UNKNOWN): like
35:19
Dave Mango: how like how would uh I mean what are you obviously there's no real answer to this, but like how do you see like this I you know, we can't just like ship a thousand things to production every day because their customers will be like, no, this is yeah
35:35
Unidentified speaker (UNKNOWN): this is
35:35
Dave Mango: not
35:35
Unidentified speaker (UNKNOWN): Yeah,
35:35
Unidentified speaker (speaker_2): I look I could I couldn't agree with you more on this data and I I don't have a great answer for this one. I mean I I I don't it it's one um
35:45
Unidentified speaker (UNKNOWN): it's
35:45
Unidentified speaker (speaker_2): one we're actively trying to figure out. I mean
35:47
Unidentified speaker (UNKNOWN): literally the
35:48
Unidentified speaker (speaker_2): meeting I had before this was like talking to somebody about this at one of one of our companies. And you know, this company I that I have in my mind is treating these things a little bit like kind of UI mock mock-ups. ups. Okay.
36:00
Unidentified speaker (UNKNOWN): You know,
36:00
Unidentified speaker (speaker_2): like they haven't really um pushed too hard on like trying to actually take stuff into production that people have developed, but they have treated it a little more as like, hey, I've got a business user, and whereas they wrote a written spec before they can give me some sort of like little interactive thing I can now understand the requirements on. I don't know
36:20
Unidentified speaker (UNKNOWN): if
36:20
Unidentified speaker (speaker_2): that feels directionally, you know, probably probably right. But look, the broader point, I think you know, I could not. agree more. Like, and this is exactly what we're working on with this company, is like trying to build a roadmap there of the tech that's connected to the business and what the business priorities
36:40
Unidentified speaker (speaker_2): are. And the company today, and this is a I mean, this is legitimately like they're grappling with this, is kind of the executives at the top of the company have this sort of feeling like. hey, we've got really smart tech people and they're producing all this stuff, but we're not really sure if it's actually helping us or or not.
36:59
Unidentified speaker (UNKNOWN): Yeah.
36:59
Unidentified speaker (speaker_2): Right. Right. And and I can see why they would feel that way. I mean, I can see what the CEO would say, like I I see them send out these emails of all the stuff they're releasing, but I don't really understand how that like moves the business forward. Yeah. Um and I could totally understand that to to be the to be the case. So you know, I I think it's I think having
37:20
Unidentified speaker (speaker_2): maybe one way to think about it would be,
37:22
Unidentified speaker (UNKNOWN): you
37:23
Unidentified speaker (speaker_2): know, even if the software itself is easier to build, uh
37:28
Unidentified speaker (UNKNOWN): the
37:28
Unidentified speaker (speaker_2): kind of old school disciplines of like getting alignment with the business and prioritizing, you know, stuff that actually creates value and bringing users along for the ride so they know how to actually use the stuff and adopt it. Like those things still apply.
37:43
Unidentified speaker (UNKNOWN): Yeah.
37:43
Unidentified speaker (speaker_2): Uh right. And even if the software you can build five times faster, great. But you still have to like kind of do the other stuff that goes alongside of it. And this particular business doesn't quite have that balance right, but we're gonna try and help them, you
37:58
Unidentified speaker (UNKNOWN): know,
37:58
Unidentified speaker (speaker_2): fix
37:58
Unidentified speaker (UNKNOWN): that. I think
37:59
Dave Mango: everyone's trying to figure it out. So yeah,
38:01
Unidentified speaker (UNKNOWN): if they
38:01
Dave Mango: don't have the balance right. that makes them normal.
38:05
Unidentified speaker (speaker_2): Yeah, yeah. That's probably true. So yeah.
38:08
Unidentified speaker (UNKNOWN): I
38:08
Dave Mango: mean, I used to one of my pieces of advice with all the engineering teams I worked with, well, I guess it's still my advice, but was always work on the things that are most important to the business. Because there'd be a lot of well, we could do this, we could do that, we could I'm like, Yeah, we we could do lots of things, but like work on the things that are most important to the business.
38:28
Dave Mango: And I think right now there's this idea that like, well, anyone can do anything as much as we want. I'm like, Okay,
38:37
Dave Mango: but like we can do lots of things, but like, does that make it a good idea? And I I think that's what a lot of people are struggling with, is that I want the dark factory, which you know, you and I could go off on a hold. whole thing about that, where anyone can sit down and write tons of software and have it come out and what and like okay,
38:57
Dave Mango: but all software is a liability. You know, you and I just
39:01
Unidentified speaker (UNKNOWN): talked
39:01
Dave Mango: about security. Like every piece of code that's out there is a potential like attack vector. And like, do I really want like my code base to explode by an order of magnitude? And but and if it's not moving the PL like who cares? Like I
39:18
Unidentified speaker (speaker_2): Yeah. Well look I I think just I agree with all that and I I would even, you know, maybe to just kind of pull that up a level, you know, I think one of the things we think about with our companies is um
39:29
Unidentified speaker (UNKNOWN): you
39:30
Unidentified speaker (speaker_2): know in some ways the technology is a little bit kind of the easy part now.
39:33
Unidentified speaker (UNKNOWN): Yeah.
39:34
Unidentified speaker (speaker_2): Yeah. You know, and so it's really the stuff back to what I was saying a moment ago, like you know, change management, getting people inside. If you have thousands of users and you develop this new technology and you roll it out, that means every single one of those users has to work in a different way and their
39:50
Unidentified speaker (UNKNOWN): job
39:50
Unidentified speaker (speaker_2): changes and so on. I I saw this study the other day that said um in this one particular area, a third of users admitted to sabotaging AI efforts at their company. companies.
40:03
Dave Mango: Wow.
40:04
Unidentified speaker (speaker_2): That's people who admitted it.
40:06
Unidentified speaker (speaker_2): Right. And so I think that's like a that's a great example of like if you roll out some new technology, it's really you have to think carefully about how to get people to adopt it, use it, and change their, you know, change their habits.
40:21
Unidentified speaker (speaker_2): And yeah, I think ultimately you gotta got to try and drive toward actually getting value uh out of it in the business. So um so yeah, I think that's all very much things we we think about. We actually think about that stuff a lot more than we think about, you know, is this model better than that model or how much does this token cost
40:40
Unidentified speaker (speaker_2): versus that token? That stuff is not really the hard part at at this point. I mean, if the harder part is all this other stuff that we're we're touching on here. So
40:50
Unidentified speaker (UNKNOWN): I
40:50
Dave Mango: think that's your your registrar story about how people were able to move to do work that they find so much more meaningful.
40:58
Unidentified speaker (speaker_2): Yeah.
40:59
Dave Mango: And like that's, you know, I think
41:03
Unidentified speaker (UNKNOWN): for
41:03
Dave Mango: some of those of us who look at the world in a certain way, like people want to go to work and do work that they feel is meaningful and valuable instead of just
41:12
Unidentified speaker (UNKNOWN): I
41:12
Dave Mango: show up for work, I do some stuff, I leave. like that's you know, we can do that. It's not particularly rewarding. Um but that's I think that's one of the things that was really great about that registrar story is it it moves people into a role that they find valuable.
41:30
Unidentified speaker (UNKNOWN): Yeah,
41:30
Unidentified speaker (speaker_2): more meaningful. Yeah, I I get it. That's that's a uh there's I like stories like that too. I mean I love those stories where
41:38
Unidentified speaker (UNKNOWN): you know
41:38
Unidentified speaker (speaker_2): that it's better for the business for sure, you know, to be have more automation and so on. But it's it's a better, it's better for the students in that scenario too, you know. So um so yeah, yeah, those are great examples. You know, look, I think people are very uh focused and excited about AI. We are, I am too. I mean, there's just like
41:58
Unidentified speaker (speaker_2): it's absolutely, you
42:00
Unidentified speaker (UNKNOWN): know,
42:00
Unidentified speaker (speaker_2): kind of a once in a lifetime, you know, experience. to live through
42:03
Unidentified speaker (UNKNOWN): this.
42:03
Unidentified speaker (speaker_2): So that part's super amazing. But look, this is going to take a generation to play out. It you know, like the AI stuff that we're seeing for companies to really change and adopt and use all this technology is it's it's gonna take years for that to happen. Button. And you know, I think that just presents
42:23
Unidentified speaker (speaker_2): for us a terrific opportunity, you know, to kind of back and find those things and really focus on helping those companies grow and scale flight technology. And that's kind of how we, you know, think about what that holds for us down the road.
42:37
Unidentified speaker (UNKNOWN): Yeah,
42:37
Dave Mango: I think I saw someone this morning posted something to the effect of like, you can go in and do all these changes, but the technology technology's is changing so fast that in six months you have to go back and reaver. reevaluate whether those changes that you made are still durable or whether you wanna do something and
42:58
Dave Mango: we're not really set up for that kind of pace of change. Like that's
43:02
Unidentified speaker (UNKNOWN): just
43:02
Dave Mango: not how we're typically have done
43:05
Unidentified speaker (UNKNOWN): things
43:05
Dave Mango: in the past.
43:05
Unidentified speaker (speaker_2): I'm sure I'll mangle it, but you know, it's the it's a little that Bill Gates um, you know, people overestimated a year and underestimated a decade, you know.
43:15
Unidentified speaker (UNKNOWN): I heard
43:15
Dave Mango: Benioff say that so many times.
43:17
Unidentified speaker (speaker_2): Yeah, yeah, I think it's very true. Yeah. So yeah.
43:22
Dave Mango: Great. Well, I really appreciate you being on the show. I'm sure Paul does too. Um that was awesome. Uh if people wanna find out more about what you're working on, what Bay Baypine Pine is working on, how how do you suggest they do that?
43:38
Unidentified speaker (speaker_2): Yeah, uh come come find me and send me an email, my name and email and everything's on the website, BayPine.com. And yeah, we'd love to hear from anyone out there who's uh got some thoughts on this stuff.
43:47
Dave Mango: Great. Um we're gonna post uh the announcement of this episode on LinkedIn. And so if people are listening and they wanna chime in on what Corey and I were talking about, please feel free to uh to join the conversation there. And thank you.
44:02
Unidentified speaker (speaker_2): Yeah. pleasure with you again.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.
All insights