Ep. 2 - The AI Squeeze PE Firms Should Run on Every SaaS Portfolio Company

Guest: Amit Pande, AI GTM & Product Advisor  

--  

AI is putting pressure on SaaS companies from both below and above, and PE firms need a sharper way to assess which portfolio companies still have durable value.

In this episode of ProPEllant, we talk with Amit Pande about how private equity and growth equity operating partners should stress test SaaS companies in the AI era. Amit argues that horizontal SaaS is getting squeezed by foundation models on one side and data platform companies on the other.

The value that remains is often deeper than the product demo. It lives in proprietary data, embedded workflows, customer relationships, and the ability to help enterprise buyers make better decisions inside messy real-world operating environments.

We also explore why AI copilots have created buyer fatigue, why “coworker” may be a better metaphor for enterprise AI, and how PE-backed companies should rethink GTM, marketing org design, professional services, and customer-backward product strategy.

Key takeaways:

• SaaS companies need to be stress tested for workflow ownership and proprietary data.
• AI is changing how PE firms should think about IT spend and human capital.
• Copilot-style AI may not be enough to create enterprise value.
• Forward-deployed services can become a product feedback loop, not just a margin drag.
• Customer-backward strategy may matter more than technology-forward positioning.

Transcript

Welcome to ProPEllant

SPEAKER_00 0:06

Welcome to Propellant, the podcast about how operating partners help companies grow inside private equity and venture capital portfolios. In each episode, we explore how experienced operators support founders, drive execution, and help companies scale. And now, here's your host, Ken Limpet.

SPEAKER_01 0:26

Welcome to Propellent, a podcast for private equity operating partners and the leaders building the next generation of SaaS companies. Each episode features conversations with operating partners, investors, and experienced operators about how value is created inside software companies before a deal, during diligence, and after the investment closes. My name is Ken Lempett, president of Austin Lawrence Group, an advertising agency that specializes in SaaS go-to-market, and your host of the Propellant Podcast. Our

Meet Amit Pande

SPEAKER_01 0:58

guest today is Amit Pandey, a product and go-to-market leader with deep experience at the intersection of AI and enterprise software. Amit most recently led new product division at C3.ai, where he ran the PL for an AI native offering and worked closely with some of the world's largest enterprises on their AI transformation journeys. He's now advising PE and venture-backed companies on AI product strategy and go-to-market, which makes him a perfect voice for what we're digging in today, how AI is reshaping enterprise value across the portfolio, from GTM efficiency to product architecture to the role of professional services. Welcome to the podcast, Ahmed.

SPEAKER_02 1:42

Thanks, Ken. Good to see you again.

SPEAKER_01 1:44

I'm really psyched to have you on our podcast. You know, when we had our pre-session, we really dug into some great and important issues that PE firms need to be considering. But before we dig into our episode, could you tell us just a little bit more about yourself and what you're up to?

SPEAKER_02 2:00

I had the opportunity to work with C3 AI over several quarters to build out new AI products for commercial execution. And one of the things I've been looking at, building upon that trajectory, Ken, is commercial execution is sophisticated and complex. And the moment you go outside of software tech, you know, into semiconductors and into the messy but interesting world of shipping to mining to all of these other industries like healthcare, that's where you're really seeing where AI is moving the needle or not moving the needle. So I'm spending my time right now with you know helping companies with really close these gaps between product and go to market, especially as AI commodifies so much. And then I'm also spending some of my time, like many others, and using Claude Code and building out small utilities for things that would have taken an army and a whole company maybe 10 years ago and now can be can be built by folks like you and me.

SPEAKER_01 2:57

Yeah, it's the whole world of everyday executives, the normal executive creating applications is kind of mind-shattering. We have a guest that was on another one of our podcasts, about probably be published the same time as this, talking about how he sees almost no value in the code itself. You know, the moat in his mindset is customers, the data that's in the system, and the people that work for the company. So it's the code itself has much less value than it did before, kind of oddly, you know, because it can almost anybody can start building things. And maybe that's a good segue into our first topic,

AI Squeeze on SaaS

SPEAKER_01 3:39

which is that there definitely is a squeeze threatening PE portfolio companies that are still stuck in this horizontal kind of SAS. The foundational models, these OpenAI, Claude, these platforms like Palantir, like from below and above, they're being threatened. So, how should operating partners be like stress testing their portfolios against that? I mean, how do you even get your mind around what the risk is to the value of the companies in a portfolio?

SPEAKER_02 4:08

Again, that's a really good way to describe it. The squeeze is real. The floor is going up and the ceiling is coming down. Very specifically, if you think about a salesperson responding to an email, a marketing person generating a basic campaign, or someone coming up with a basic spreadsheet that calculates terminal value, those things are getting easier to do with foundation models. And that's your horizontal SaaS doing a regular, you know, person's sort of horizontal job, if you will. And then the unified data foundation story by the paliantiers of the world, the Databricks, C3 and companies like that, while nobody wakes up in the morning and says, I want a unified data foundation, whichever company has that unified data foundation is in a pretty good position to then add agents or applications on top of it versus sometimes the other way around. So I think the interesting question is the squeeze will continue, but what is valuable in the middle? And so if I think about private equity stress testing their portfolio,

Defensibility Data Workflow

SPEAKER_02 5:04

it goes back to something you framed our question around, which is the value in the relationships and the data in the workflow is where I would start. Look, most PEs acquire companies when you know things are a little bit funky. The company is not necessarily in its prime. And so they get it at a very fair valuation. And often, if you look at these companies from the outside, if you look at their marketing, if you look at the way their product is represented, you might mistakenly feel whether you're outside or you're in PE, you might mistakenly feel that, oh my God, this company is not defensible. But that's just the surface. I think you want to go three clicks below the surface and say, what is the workflow that is embedded today? That is in the day in the life of an end user, whether it's in the realm of immunology, in the realm of shipping, whether it's in the realm of mining, it could be any such industry. If you own a workflow with your software in the day in the life of that end user, I would start at that point and say, are we owning a part of the workflow with proprietary data or something that we're enabling in terms of a better decision that the foundation models can't do because they don't have access to the data? Or then the platform companies can't do because you don't have to build everything on top of them. I would start there. If the answer is actually no, in this stress test, we're not finding anything that is a sticky decision, a deep relationship, or some kind of remote, then I think you have a problem. You have to think about how to then take that horizontal business and do something with it. But if the answer is yes, then I would double down on that data and workflow. And then that opens up a new conversation on, you know, how do you take that 2x or 3x multiple into something that is several times five?

SPEAKER_01 6:48

Yeah, you know, everything seems to keep coming back to those two things, data and workflow. The workflow representing the knowledge of the corporation, right? Its ability to execute and the data, especially if it's proprietary, being this unique advantage it can bring to decision making or acting in the market. So I actually wrote a Substack on Moat, the title is Motes Make Goats. And one of the main motes I wrote about is data, because it's been since my beginning of my career in technology, which I'm afraid to admit, but goes back to 1980, the data has always been the thing that was the advantage. You know, the compute, the logic of a piece of software might have been something of a barrier, but I wouldn't ever have called it a moat. So I think having business logic, business intelligence that is well understood and a proprietary data advantage can't be underestimated. And that's, I think, as you just said, where to look for, you know, are these companies going to have survivability?

Beyond Copilot Mediocrity

SPEAKER_01 7:51

When we talked before, you you mentioned that the Microsoft co-pilot experience has sort of conditioned enterprise buyers to expect mediocrity, right? And that anything that's kind of chatbot first can sort of take all the air out of a room, right? Sort of kills kills the vibe. So, how should portfolio companies reframe their AI narrative so that they don't fall victim to this kind of buyer skepticism?

SPEAKER_02 8:19

It's a great question as it relates to, in fact, what you spoke about, you know, with uh institutional memory and context, right? Because you you sort of have this structured, captured context of an enterprise that is in all these enterprise systems that you and I have seen in our different generations. And then you have all of this implicit context that's in people's heads and it's in the relationships and the things that never really get documented. And then there's all this external context outside the cooperation of its context, you know, if you go back to kind of, you know, port of five forces, right? All of this is the context that AI can actually feed on. And today we're dealing with a very small slice of that. And the proprietariness of the data that you get with all these other contexts make it interesting. As it relates to Copilot, Copilot certainly works in only some of these silos, right? And so part of where the copilot era has now become a bit of an exhausted metaphor is that you're able to ask it questions about things that are often just a little harder to look up. Now, I will say that we should acknowledge what copilot has done well, which is that it has, for many of us who have been waiting for the personal AI assistance and Siri promised us this in 2011, it actually showed us at least some of those assistant capabilities, you know, while GPT and Cloud caught up. You know, but the numbers that you and I and both know can are not great, right? You have 6% paid users in a 450 million Microsoft offer subscriber base, and you're making five or six billion on Copilot, which for Microsoft is, you know, one-fifth of what they spend on data centers every quarter. So we have to thank Copilot for doing phase one and then stop there and say, this was a great diagnostic. You've shown us the direction, but now let's move on. And in terms of where we're moving, what becomes interesting is going from either the metaphor

AI as Coworker Metaphor

SPEAKER_02 10:06

of co-pilot to that of a coworker or the metaphor of collaborative intelligence. And this is where you really start going into the harder question of is this AI paradigm going to help me make better decisions in my workflow versus just answer questions for me? And so I'll say that, you know, when I think about the coworker metaphor, I know it's a bit of a lightning rod today because, you know, are these coworkers that are AI going to become the co-workers that power an entire cooperation? But I think what we should take from there is that when we work with exceptional people, we expect something from them that is a level of intelligence or insight or an action that we can trust them to do. And I think as we move towards these models of trust anchored with proprietary data, you know, that's where I think we are going to move into newer metaphors. You know, at the end of the day, the end user cares less about the co-pilot. The procurement manager cares more about did I get my prioritization on budgets done better. And the person that's a field operator cares more about did I identify an anomaly that is going to fix this soil pipeline in the field? And I think that's where you can directly start connecting this back to something that is either margin or revenue or profitability or cost reduction. And, you know, and I think that's where we'll see co-pilot 2.2.0 or whatever we call it be successful. I actually think this is how we'll get out of the hell of pilots that you know we're all aware exists around us today.

SPEAKER_01 11:36

Yeah, I love the idea of the coworker rather than the, you know, the chat buddy to have an array of coworkers who could help me in my work, whether they're virtual or human, I think would be very valuable. I saw, I can't remember now the name of this remote working tool where they have sort of your whole organization online and you can sort of tap into anybody at one time. And it's like a perfect place to be putting both our, you know, our human coworkers and sort of the knowledge workers that the AI can deliver, you know, in one kind of uniform place. And I think that that would be a great direction for operating partners at PEs to try and drive their portfolio companies. How can you move beyond, you know, this co-pilot? And I think you've framed it really well, save me a search step resource to something I can interact with that's actually going to make my job a little better. And I think if I understand correctly, the role of these operating partners, they have to kind of drive thinking within their portfolios, right? So that might be a good place for operating partners to drive some thinking. How do we get past that first phase? Kind of digging in more on this the coworker kind of question. We talked about the idea of the first three roles you'd hire in marketing when we did our prep session. And you suggested that you'd probably answer

Rebuilding Marketing Org

SPEAKER_01 13:01

the question, you know, what are the first three roles you'd hire very differently today than just even two years ago. So walk us through kind of your thinking about what is for a young company, you know, what does their marketing department look like today versus what it might have looked like three or four years from now? And what does that mean for the PE operating partner?

SPEAKER_02 13:23

The co-pilot era that we just covered was really around bringing some AI seasoning, if you will, on the human pyramid. You know, we've had the human pyramid since the days of Taylorism, maybe in the 1950s, right? It's been it's been an org structure, is what it's called, I guess. But really, where we are moving to is and what affects the organization design or redesign is that it's almost a distributed intelligence network, right? I love the visual you painted of, you know, you can actually see your entire org of humans and AIs together. And so now if you take this in the context of the go to market, let's talk about marketing specifically. There were some very simplistic notions in the past because of the way boundary conditions evolved. And we got to this point that you had a demand generation person, a digital marketing person, a product marketing person, an operations person, and then you had sub-specializations. Like, you know, is there's a technical SEO person, and and I can tell you that every single CMO that I've worked with, and I've been a CMO also in the past, you know, we would think in those terms when we spoke to our boards. And the investors on the other side also knew that, well, if you need a full-stack marketing team, you need these like 10 functions. Why? Because if you needed more content, more campaigns, and more coverage, you just needed more people. What AI changes fundamentally is that you can now deploy a much larger coverage campaigns and content playbook. And so now you don't have to worry about the limits of that scaling. You actually have to worry about the harder question, which is do I really understand my bias journey? Do I really understand where they are and where I can meet them? And so the way you now redesign these organizations is if you're truly workward working backwards from the bias journey, you think about a few very high-leverage marketing operators who know how to wield the power of agents around them. And I would go so far as to say that, you know, in the in the 80s or 90s, can Isaiah Singer wrote this book called The Hedgehog and the Fox, right? Where he said, you know, one big idea versus several ideas. You have to be a hedgehog in something. But I think this is the era where you need to be a fox in several things. And if you have three people like that, and those three people might be the orchestrator that brings all these sort of agents together, and that becomes the future of marketing ops, that person is the storyteller. That in a sense is your corporate and product marketing person like together. And then I do think that when you get exhausted with digital, as we all are, there's nothing like the power of real events. You and I recorded the SaaS Backwards podcast at an event. You know, we'll remember that event. And so I think these roles, if you expand outside of marketing, think about sales, think about product, think about finance, GNA, all of these can be reimagined. And what this means is that it means two things. Private equity partners I work with tell me that the top two things on my mind across my portcodes are IT spend and human capital spend. And so you can address both of those through a new kind of organization design model. And if you're a private equity partner today, looking at your organizations, I think you really want to think about how to break down the current jobs being done into jobs to be done. And then you can do the exercise on the repeatable aspects that should be done by AI, judgment, repairing trust, rebuilding trust or establishing trust in a new geography, a new vertical, and all these nuances of customer interaction. I think there's new jobs that will be created with AI as well. So I'm actually bullish about a smaller set of people doing more, but perhaps getting rewarded more as well with AI. I think it's a win-win model.

SPEAKER_01 16:58

Yeah, you know, I want to call our listeners' attention to the LinkedIn presence of a client of ours and also couple time guest on our other podcast, SAS Backwards. His name is David Gabriel, and he's building

Agentic Teams in Practice

SPEAKER_01 17:12

an entire agentic product marketing team out in the open. And so on his LinkedIn, and again, he's David Gabriel, and the company he's with is called Rumbics, R-H-U-M-B-I-X, fascinating build-in public, agentic product marketing team. And he described exactly what you have. He's got himself and two others, and they are orchestrating the marketing with the help of these agents. And he described the hardest part actually is being able to delegate to that many effective players. You know, his span and scope are much greater as he's built out this agentic organization. And he actually has agents managing agents. So I just think it's a really interesting place for people if they want to dig in on what does this actually look like in the real world. It's not just us spouting off here on a podcast. It's actually happening.

SPEAKER_02 18:06

And Ken, I'll uh I'll close with one more comment on that question that there's a lot of simplistic assumptions going around today, right? Let's just reduce the whole marketing team down to one person. Or, you know, this is really more of a headcount planning decision. No, this is a capital allocation and leverage design decision. That's how you want to think about it. Because you're trying to design the organization that will also sustain itself in the future. You bring your marketing town team down to zero, I would argue you have a bigger problem than if you bring your sales team down to zero because you've suddenly lost all this context and then you're being pennywise, pound foolish. And I think when I talk to the best uh private equity partners, I'm definitely hearing that they're starting with what you mentioned about David's work, which is you need to understand your existing process really well and map it out before you can even start bringing in more and more agents to do the work. Now you can do more with them versus a hundred percent UN organization. Totally.

SPEAKER_01 19:02

Hey, I want to keep moving because we're gonna try and pack a lot into this episode. For a long time, professional services has been sort of the a detriment toward the software multiples, right? The multiples of these companies. But in our prep session, you made the case, or maybe we did it together, that forward-deployed

Services as AI Advantage

SPEAKER_01 19:20

engineers, that model might actually be an important mechanism for unlocking AI value in enterprise customers. How do you see PE firms reconsidering professional services as a strategic asset and not just a drag on the multiple?

SPEAKER_02 19:37

There's a good reason that private equity was and venture capital was wary of services-heavy models in the past, because you had engagements that would not generate any IP or any repeated intelligence loops that would feed back to your product or your business. You essentially were doing resource for hire type of models. What has changed in the forward-deployed model, which Ballantry deserves a lot of credit for, is that you embed really smart people to understand the messiness of somebody's world. And I can tell you, Ken, from my experience in working with some of the largest Fortune 500 companies in these gnarly industries, okay, biopharma, semiconductors, and such, that it is 10 times messier than you think before you start working with them. We're talking about 500 data sources and maybe five years of dozens of Power BI tools people have built on an older generation of ML models, a process that involves people in Austin and in Houston and somewhere in Penang, Malaysia. And the people in Penang, Malaysia only work on Excel sheets. Forward-deployed engineers and that model of next generation AI services, it does three things really well. And that's why private equity is paying more attention to it. First, you embed and start with understanding the customer's mess rather than starting with the technical elegance of your solution. I think that means that you're reducing buyer fatigue. And buyers are saying, oh my God, you actually understand me in a very therapeutic sense. You're here to really understand my messiness and not judge me for it or try to shower technology. Number two, what you're doing is you're creating this intelligent product feedback loop. Because when you're capturing someone's domain into a data model or an ontology, you're creating these repeatable black boxes of intelligence. That then can feed back into your product and help you determine if you should now build new agent layers or you should build new accelerators or units that can then feed back into your product. And in the domain or jurisdiction in which you operate, every next customer you go to, now you actually have new pods you can add to your earlier product. And I think that's why this model of services not being a drag, but being an intelligent loop to the product you should become, which you can price higher. I think that is now possible in a different way in the AI era. I hope we can dig into this more someday in the future as you and I learn more about companies that are deploying the palantry model without paying. A million dollars a year for the parent DFTE, which is one of the reasons why it hasn't scaled as much.

SPEAKER_01 22:04

I really like the way you described it. I think there's a huge opportunity. And we used to do this a long time ago in IT or surrounding software products. The customer would pay for the development and it would become part of the product. That's, you know, even before cloud was a thing, that was how those additions to the software got funded in many cases. So it's almost like we're going back to the future here, but in a probably a more powerful and quickly monetized way. So it's actually very cool. And in fact, one of the things we talked about, and we don't often talk about professional services firms in our podcast, but I think it's important that we just kind of make a note here about the opportunity for professional services firms to get into this business. A long time ago, we worked with Cognizant as it was addressing the needs, the perceived needs really of Y2K. And it helped that firm grow up from being truly a body shop to project capable and ultimately really a major player. It was one of the one of the real launching points for the Cognizant we know today. And I just think this is a great opportunity for firms to dig deep into customer workflows, sit with these users, and actually build presence and credibility in a way that they haven't before, and maybe stem the job loss and business loss that AI has already wrought on them. And I'm wondering if you share my enthusiasm for the possibilities for the professional services world. 100%.

SPEAKER_02 23:35

I have seen how Cognizant and Accenture, among all the things that they did well, one of the areas they went deep into is design thinking. Design thinking was starting to come into our corporate consciousness maybe two decades ago. And by investing in design firms, Accenture, I remember, actually acquired half a dozen design firms around the country. Deloitte did the same. Cognizant was the most forward-leaning of the India-based companies at the time. That means you have an ethnographic level empathy for the mess that exists in a customer, but you can capture that in a meaningful way. And then that becomes encoded back into the offering that you can do. The only other thing I'll add to your comment, Ken, 100% aligned on what you said, is trust is owned by whoever's closest to a customer. And one of the things these firms have been great at is they own the trust with customers in many industries in a way that for today's world is whether it's corporations or geopolitics, we're in a world of bridges and moats and fortresses again. And you need insiders who know how to speak to insiders to even earn a conversation with someone that, you know, let's say works in a European pharmaceutical company because they're actually scared about words like disruption. And I think this is where professional services firms also have the advantage that if they bring these forward-deployed engineers and combine them with actual AI builders, there's nothing that stops these companies from generating a parallel product line and monetizing that and having a product forward business also as part of them. I'm actually hoping to see some successes come out of this in the near future. And I think they'd be good companies to invest in.

SPEAKER_01 25:16

Yeah, I mean, these firms have shown an appetite for building applications in the past too. So it's sort of an interesting possibility of the professional services firms becoming, you know, AI powerhouses in their own right. It's kind of interesting. Kind of the last thing I want to touch on is that we talked about, you know, winning vendors aren't necessarily the ones that are going to have better models. They're firms that and through their forward-deployed engineers, kind of walk into a client that might

Customer Backwards Selling

SPEAKER_01 25:46

have this crazy stack of legacy machine learning, which I think you talked about just a few minutes ago. The scattered SAS, you know, all kinds of things that are sort of orphaned and say, let's start where you are. And I think it might be good to land our episode on the idea of what is a customer backwards approach, which I think you were starting to describe, and how would that be the technology forward pitch, and what does that mean for our operating partner listeners?

SPEAKER_02 26:16

The technology forward pitch worked for a certain era because the technology was incredibly superior to whatever came before it. The first Snowflake instance, the first Salesforce instance, the first time you and I probably had an iPhone, it was materially superior to whatever existed before that. That era is gone now. Technology exists around us in a way that if you now lead with technology forward approach, the problem is you're really alienating the buyer. You're getting them scared about this really expensive future that is going to be disruptive, where they're already hurting and a little bit fatigued. And so the number one mistake that I see a lot of technology companies here in Silicon Valley do, which they have to stop doing, is to not talk about how great your platform is and you know, stop showing one of these things that is an eyesore. The customer backward approach really starts with first, often when you document what a customer's paints are and what their existing day in the life is, you're actually doing the customer service scan because the customer hasn't had the time to actually document this. I saw in one of my most successful deployments last year, it was a semiconductor customer. We went from level one, level two, level three of actual process mapping. And I'm talking about the hard work of process mapping together with your customer. And their chief business officer appreciated it because they said, you know, we've never done this and we're a public company. And now you have helped us build something that we can use for a foundation for any application we build into the future. But second, the question of where agents come in and start helping decisions in a human workflow, where the exceptions are to those decisions, what becomes repeatable, if you're a product team, you're very distant from that messy world. And we have to reduce that gap and reduce that distance with this customer backward approach. If you do it right, you're actually connecting customer pain and new product strategy with what your new product marketing should be. And if you can get that sequence right, Ken, well, that's the difference between a business that is seen as, oh, it's kind of this vertical or horizontal business from 10 years ago or 15 years ago, which is what a lot of PEs have. To wait a minute, this is a business that is literally mission critical to the industry that they serve. And so I think connecting this back to role design, I would actually say that I see a lot more roles emerge in the future, even for new grads, at this customer interface. And I can tell you, I have I have leaders in my network who have told me that I've been a CEO, I've been a CRO, I've been all this. I want my next job to be like a customer operator. And I'm like, what do you mean by that? And they're like, I want to just spend a lot of time working embedded inside a customer and then work backwards from there to see what we can build. So you're right, it is back to the future because we're going back to a time when, you know, there was a time when you had to pay a lot of money for an expensive bespoke suit, right? And then, of course, all this era of new, you know, fashion came in and you know, bespoke spoke becomes really expensive. In software, we're actually going back to that era where you don't have to worry about buy and build, you will actually buy with the ability to build on top of it. I actually think that the vendors that are embracing this level of interoperability and letting their customer choose the Switzerland they want to play in, they're going much further than the ones who say, if you standardize everything on me, you'll win because buyers are just much more wary of that. And they want the optionality to, you know, play with this like Legos in the future, because that's how you can change the org model and the technology layer at the same time.

SPEAKER_01 29:52

And that's a great place to land our episode, the idea of hybridization of the software, the buy-build idea not being as relevant as it has been for so long. Hey, Amit, if folks want to reach you, learn more about what you're doing, and possibly engage with you, how can they find you?

SPEAKER_02 30:08

LinkedIn's the best place. And I look forward to us continuing this discussion that we started with SaaS backwards several years ago. And very excited for the work that you're doing for the investor community.

SPEAKER_01 30:20

Thanks so very much. And if folks want

Wrap Up and Contacts

SPEAKER_01 30:22

to reach me, I'm on LinkedIn slash in slash Ken Lempit. My advertising and SaaS go-to-market advisory agency is called Austin Lawrence Group. We're at AustinLawrence.com. And the Propellant Podcast is now available pretty much wherever podcasts are distributed. If you like this episode, please consider subscribing. Hey Amit, thanks so much for joining us today on Propellant.

SPEAKER_00 30:45

Thank you so much, Ken. It was a pleasure. Thanks for listening to the Propellant Podcast, where we explore how operating partners help companies grow inside venture capital and private equity portfolios. If you found the conversation valuable, please subscribe and share the episode with another operator, founder, or investor. And we'll see you on the next episode.