July 23, 2025

Simon Walker: Forget 'AI Strategy' How PE Firms Can Drive Real AI Impact Across Portfolios

Simon Walker: Forget 'AI Strategy' How PE Firms Can Drive Real AI Impact Across Portfolios
Simon Walker: Forget 'AI Strategy' How PE Firms Can Drive Real AI Impact Across Portfolios
AI Pathfinder for Private Equity Podcast
Simon Walker: Forget 'AI Strategy' How PE Firms Can Drive Real AI Impact Across Portfolios

Summary In this episode of the AI Pathfinder for Private Equity podcast, Steve Budd speaks with Simon Walker, co-founder of Kubrick Group, about the critical role of AI in private equity and portfolio companies. They discuss the evolution of AI capabilities, the urgency for private equity firms to adopt AI, and the importance of collaboration among portfolio companies. Simon shares insights on real-world applications of AI in sales and finance, the necessity of measuring AI success, and the c...

Summary
In this episode of the AI Pathfinder for Private Equity podcast, Steve Budd speaks with Simon Walker, co-founder of Kubrick Group, about the critical role of AI in private equity and portfolio companies. They discuss the evolution of AI capabilities, the urgency for private equity firms to adopt AI, and the importance of collaboration among portfolio companies. Simon shares insights on real-world applications of AI in sales and finance, the necessity of measuring AI success, and the cultural challenges faced in AI adoption. The conversation emphasises the need for proactive strategies and clear ownership of AI initiatives to drive value and innovation in the private equity sector.

Takeaways

  • AI is becoming essential for competitive advantage in private equity
  • The urgency for private equity firms to adopt AI is increasing
  • Collaboration among portfolio companies can accelerate AI adoption
  • AI initiatives should have clear ownership to avoid failure
  • AI should be viewed as a tool for tactical wins
  • Setting realistic expectations for AI outcomes is important
  • AI products require ongoing investment and development

Chapters
00:00 Introduction to AI in Private Equity
00:57 The Evolution of Kubrick and AI Solutions
02:50 The Urgency of AI in Portfolio Companies
04:32 Use Cases and Real-World Applications of AI
06:13 Mandating AI Implementation in Portfolio Companies
08:40 Accelerating AI Adoption Across Portfolio Companies
11:55 Measuring AI Success and Setting Expectations
13:47 The Importance of Defined AI Use Cases
15:29 Standardizing AI Readiness Across Portfolios
16:18 Identifying Skills Gaps in AI Implementation
19:23 The Role of Stakeholders in AI Projects
21:41 Kubrick's Focus on Talent and Capability
23:46 Preparing for the Future of AI in Private Equity
25:42 Cultural Challenges in AI Adoption
29:13 Conclusion and Key Takeaways

Links

Simon's LinkedIn: https://www.linkedin.com/in/simon-walker-8763b9/

Kubrick Group: https://www.kubrickgroup.com/uk/

AI Pathfinder Private Equity AI Strategy Meet-ups:

https://www.aipathfinder.co/private-equity-ai-events

Have a question or would like to suggest improvements, please contact steve@aipathfinder.co

SPEAKER_00

Hello and welcome to the AI Pathfinder for Private Equity podcast. I'm on a mission to help private equity move from AI curiosity to confident capability. And one of the fastest ways to learn what really works is by speaking with founders who are building focused solutions and services that tackle PE and portfolio problems. Today we're looking at the critical moment when a portfolio company begins to build real AI capability, not to experiment, but to deliver real value. And we're exploring that and what that means in practice. Joining me is Simon Walker, co-founder and chief commercial officer at Kubrick Group, who worked closely with private equity and portcos to build AI products and capability. Simon, great to have you here. Yeah, thank you for having me. Well look, can you kick off with a quick intro on yourself, how you came to found Kubrick, and probably a little bit on the critical problem you're aiming to solve now and in the future for P. Yeah, sure.

SPEAKER_01

So we set Kubrick up back in 2016, and the original premise of Kubrick was we saw organizations had the capability of collecting all this data at the time. I don't know if you remember, but everyone was calling it big data. It's a term you don't hear too much now. But um organizations suddenly uh realized they had the capability to capture huge amounts of data and the technology to analyze it, therefore gaining very, very valuable insights. The issue they had was that this was a this was a people problem. So they had all the technology there, but they didn't have the people, they didn't have the data engineers to correlate and organize the data, they didn't have the machine learning engineers to be able to provide frameworks to analyze that data. So we set Kubrick up to answer that. So we saw that uh society and organizations had a digital skills emergency. And it was really important to make sure not only those individuals that were coming in, so we created the workforce, but it was also a diverse workforce. So you had a diverse set of questions being asked of that data. And really we've transformed, but the problem has only slightly shifted. Post-COVID, I'm quite sure that a number of people listening remembers that uh just towards the end of COVID, the consultancies received a uh significant boost and wind in their sales as everyone rushed to digitize. We came to the end of that. Um, really, suddenly the C-suite focus was on AI. And the problem shifted from being necessarily a digital skills emergency. And actually, what it shifted to be was how are we going to get a return on the investment of all these hundreds of millions of pounds we've spent at centralized data teams? And that's where Kubrick entered its next phase. We build AI solutions that solve real business problems, but we also build very, very capable workforces with uh tomorrow's skills.

SPEAKER_00

Great. Yeah, thanks. Thanks very much for that. I think that um we're we're looking at quite an interesting dynamic in in this conversation, aren't we, about the the private equity and their portfolio, that they are for you know, for a number of years, they're the guardians, if you like. Is is there then a sense of growing urgency for them to be more hands-on when it comes to identifying replicating value um AI across across the Yeah, I mean for me it is.

SPEAKER_01

So reference one of the reports that's recently come out by Bain that if you look at sales or if you look at sales divisions within companies, that they're getting a 30% improvement um in sales win rate by implementing AI. And if you're a a private equity company with a series of portfolio companies, you're not gonna be wanting to have a sales team, for example, that isn't utilising AI, because if your competitor with a very similar product is, then you can see that you're you're you're gonna be behind. I think I think the time is now, and also the barrier of entry to cut to be able to access AI is really low. The tools are there ready. The focus and investment has to be in how you can get a return on that. But yeah, it it is critical that the private equity companies are expecting AI to be implemented and used throughout their port codes.

SPEAKER_00

I think that initially, and I'm sure still is the case, my network's a year in, and at the very beginning, it was it was purely helping P to understand the value of AI. We probably weren't at the stage yet where they were able to have um they'll probably shoot me for this, but competent, really competent conversations about the application of the technology. That's that's definitely changed in a short period of time. You talked about one example there around sales. What other use cases are you seeing real traction?

SPEAKER_01

So, I mean, well, sales, sales is is is is a good place to start because there's new numerous elements on there. And I'll come I'll come back to those actually. But um, if you look at finance divisions, so um FPNA, uh talking about the barrier of entry, uh co-pilot co-pilot, I would say, is ubiquitous across a number of organizations and a number of uh across support codes. Um, but actually getting chat GPT style of agents build rolling forecasts, you know, automate anomaly detection, which frees up finance teams to do far more value adding and strategic work. There are there are some uh private equity companies out there, and they've been very public about their companies have to have provable AI capability and driving certain business processes. So I know of an example where um VistaRons uh is one of their portcos has built a relatively simple sales tool, but is super effective, where Copilot is a harvesting SharePoint. And what it's harvesting for is it's looking for examples of where they have done certain things with certain clients, what was the business, what was the value add. It's creating automated emails. It can reply back to certain emails, freeing the salespeople's time to do other things. And that that that front office sales element uh suits very well to very simple, in this example, co-pilot examples where you can implement them in and you can increase the productivity of the sales team.

SPEAKER_00

Oh, you talked there about MP being insistent. Is that an insistence with their portfolio that those sorts of tools have to be implemented? It's almost mandated.

SPEAKER_01

Yeah, in um the two examples that I that, again, as I said, that they're out there and they've spoken about quite a lot is Apollo and Vista, and they have mandated that their portcodes have to be able to prove out AI use cases and that it's and it's generating productivity or it's making them more resilient or it's saving them money. I mean, the interesting thing I I think, if I was a private equity company at the moment, is what happens when you come to deal time and you don't have any AI capability. And I think there is an expectation from buyers that they would expect to see AI being used through the company. Now, it might not be the finished article, depending where they are, but you would expect that to see because isn't it a measure of the management team's capability and the management team's, I suppose, horizon scanning or being ready for the next best thing? And I think if you don't have it, you run the risk of significant price chipping.

SPEAKER_00

And you are a portfolio company, and well, obviously you have uh quite a bit of capability there, but are you getting that sense? Has there been a change of sort of tone from your from your owners?

SPEAKER_01

Yeah, definitely. I mean, when when it was interesting when we were a few weeks ago at your great event, I think that a very rough straw poll and show of hands was a third of companies felt that they had examples of AI being used in their portcos and being used, or a third of their portcos, I should say, were using AI and using it well, and there was some they could evidence some uh productivity gains and efficiency gains in that. Um, another third were kind of on the very beginnings of a journey, and another third didn't know where to start, and and that was across most of the private equity companies, and that was a reflection of their portfolio. So, what I got from that was that the awareness has definitely increased. I mean, if you took that 12 months ago, the awareness has definitely increased. And I think with awareness comes expectation, and you know, for the third that really aren't doing anything at all, I think it's critical for them to be able to start exploring what they can do with those companies and where are the where are the elements of the business that suit well to having AI tools within them.

SPEAKER_00

Yeah, well, let's get into maybe some examples. Certainly, how how should P apart from mandating, are there ways that they can start to interact with each other, the portfolio companies? Because there's going to be those that are further ahead than others. And in the conversations I've had, you know, that there is some actually seems to happen more on the VC side, in fact, where there's a little bit more of a community there where they're they're sharing quite a bit. But yeah, what are the ways that they can sort of accelerate this within their, I don't know, 30, 40, 50 companies?

SPEAKER_01

Well, I think I think you touched upon it that actually having um a community of your port codes together is a really good place to start. So often they're the port codes are non-competing, often, so therefore um there'll be a far greater willingness to share things. I was a while ago, I was at a uh CRO conference um run by Bomark, and it had, I think, eight other organizations there, eight other CROs, and a great chunk of the day was talking about AI tooling from uh from a sales lens, and I found that really useful. There was an element there around campaigning, outbound campaigning, that was and I'd never thought about it like this. It was from a completely different business, but very applicable to ours. So I think you would you would want to start with kind of that ideation of your um of your group of port codes, but also actually understanding where actually having an AI scan across all your port codes and understanding where they're at. And um I would be asking them, and I think this is really important actually to differentiate, I'd be asking them what what other things they think they can apply AI to, not what their AI strategy is. So I think once you start talking about an AI strategy, it becomes kind of all-encompassing and quite um overwhelming. And actually, AI is a tool, it's tactics. So you can, if you're starting on your journey, you can pick two or three use cases that are well suited to AI, and and hopefully other portcodes of your uh investor will have some of those ideas for you. And I think I'd be asking them what's the hundred-day plan around those, and get everyone to start presenting back. Just like I felt that PE companies, you know, going back three or four years, they were they were driving on what's your plan around DEN and I. It's exactly the same. I'd like to be seeing that from an AI perspective. So, what are the plans? What have you got? What's the measurements that you're expected to have on these? So, you know, there's really only three things it can do. It's either gonna drive revenue, it's either gonna make you more resilient, make you um more easily fall in line with the regulation or mitigate some risk, or it's gonna be more cost effective or make you more efficient. So I'd be wanting to know what do they think they're gonna get those, what are the measurements, and I'd be I'd be wanting to measure them every quarter, or in or putting a uh side in the monthly boards, which that other element I haven't seen too much evidence of, but I'd want to be seeing a monthly report on where am I with the one or two AI initiatives that I'm running through the business.

SPEAKER_00

Talk to uh a couple of things there, perhaps we could dig a little bit more into some sort of benchmarking or standardized scoring across the port codes, but also this measurement piece. And it can often, I think there's uh you know, conversations I've been in is that we're we're using AI, but I I'm not feeling any benefit yet. Yeah. It's that so you know the people are applying it but aren't yet feeling what improvements it's making. So, you know, having some form of data point is really important. Could you talk across those those two things in a little bit more detail?

SPEAKER_01

Yeah, well, I think one of the things is on the point that you made, I'm not feeling any kind of return or benefit from it. I do think you should also start with AI because some of it is experimental, but it should be very low-cost, low-time experiments. And there has to be a no-go point. So go no-go point. If you are not making the return that you thought, or it's not hitting them, whether it's a a time-saving metric, a usage metric, or uh, you know, an e-bit dar uplift, if it's not hitting those over the time that you've said, I don't think people should be scared to drop it because they should be controlled explosions, but they should be low cost. So often if uh if a client came to me and said, we are not getting the return that we expected, I think my first point of call would be: has it been defined correctly? Were the expectations set correctly? You know, what are you looking at? Has it been defined what you're trying to aim for? And is it realistic? If those are all yes and it's still not giving the return, well, probably wasn't ever going to give the return. So I think there's sometimes a bit of fear of failure with this. Although it's uh it's probably well overused, the fail fast comment on this is is, I think, very accurate because it is experimental. Um and you're not always going to uh you're not always going to get a return immediately.

SPEAKER_00

Yeah, I I agree. I think they're probably setting some sort of expectation that you know a certain percentage is going to fail. Be be okay with that.

SPEAKER_01

Yeah, I would like to see if if I was on the other side, and obviously not I'm a I'm a portco, but I would want to see from a portfolio company, and granted there's various sizes, but let's say and we are so mid-size um organization, I I would like to see two or three AI use cases that are well defined, that are measured monthly, and have a have a have agreement from my investors what we're going to do, and also have be very clear on the go-no-go's and report on those. And don't worry about if one or two of them fail, because not all of them are going to fail. You know, you will out there get get the winners. And once once you start succeeding on one, then that's probably the time to um invest more in the proprietary data that's going in and the frameworks that are going in, because that's when you get real return.

SPEAKER_00

As a uh product manager, past product manager, having sort of kill metrics is crucial, especially when in the past the work for founders that everything's sort of a pet sort of project and wants to see it uh grow. Um, you know, it's their baby. But yeah, having some kill metrics.

SPEAKER_01

Especially in this, you don't you don't suddenly want people just forcing it through or any vanity projects. This isn't this isn't, you know, AI tooling and experimenting isn't good for that.

SPEAKER_00

Well, the scoring piece, the standardized scoring, is that about you know, across the port codes? Obviously that there's it's really important for the P firm to have a good visibility, but is there is there some other benefit, gamification, is there some other way that they could compete or you know, in a f in a more sort of I don't know, yeah, not fun, but I go back to um I go back to when uh DEI was being driven through portcodes.

SPEAKER_01

Um there were quite there were transparent scoring about around how everyone was doing, what what you know, there was uh rag reports on it and what needed improvement and what needed focus. So I think the same gamification as you said, or standardization and reporting should be used in AI and AI readiness.

SPEAKER_00

Yeah, there's probably worth as moving moving on, I suppose, into literacy and and capability, because no doubt in those measures you're gonna find gaps, aren't you, in in being able to perform, I suppose, scaling potentially these opportunities. What what are those skills that are tend to be missing within the portfolio to to achieve that?

SPEAKER_01

Yeah, I mean this is this is a across organizations as well, so it wouldn't be just portfolios. So what we're seeing at the moment is that uh if I go back to originally where Kubrick set up, I think one of the issues that building out large data teams had is they often felt guilty of not focusing on what was the business problem they were trying to solve, kind of became tech for tech's sake. So they picked up a data handbook, they went, right, we've got to drive, we've got to build a data lake, uh, we've got to get data quality like this. And there was, there was no what is the business reason that we're doing this. So my first point of call would be looking at, you know, is it is it monetization we're looking for? Is it efficiency we're looking for? What are those things as the business goal? So you don't start getting technology teams doing tech for tech's sake, and there is a business owner around those that is a stakeholder on that is the output that I want. So to answer your question, you need individual that that have is a bit of an analytics translator. Um, you know, they they understand prompt engineering well and how to do that, they get the domain, they get the business challenges, and they can connect all those dots together. And I think those are the individuals missing because you can certainly fall into line of having this is our IT team, this is our data team, this is the business unit. And AI does not work well in siloed businesses, and you need individuals to own it. So, you know, my my big thing is looking at data teams that have failed previously, it's because they are in silos, different silos, and they haven't actually communicated and understand the business goal.

SPEAKER_00

Exactly the same hazards um you have for uh well, I think those closest to the use case, you know, the workflow that they've been potentially running for years and years and years. We held a session a couple of quarters ago where somebody created a cookbook for the analyst sort of role. And so it was really breaking down all of what they they were doing and then seeing where AI could add value to that process. But it's really those that are doing that role can really understand and do that. And then I and then I I think there's a uh I think it's the trying to think of the word now, but it's being able to review and know that there's a successful output or not.

SPEAKER_01

Yeah, yeah, uh, yeah, I totally, I totally agree. I mean AI um AI pay AI pilots fail really without the proper stakeholder buy-in. And you really do need someone that can translate those outputs into actions, you know, whether it's investor oversight, compliance risk mitigation, sales lift, clear OI, whatever it is. So you need someone to pull all those together and to and to manage the stakeholders around those. It can't be a side-desk project for someone to pick up and put down.

SPEAKER_00

No, a judgment was the word I was looking for. Is you know, someone that's able to make that judgment. Yeah. You know, the human human, the qualified human in the loop, if you like.

SPEAKER_01

Yeah, exactly. Correct.

SPEAKER_00

So how important are the sort of commercial and I mean we've you've talked a little bit there about commercial, um, the the the comms school as well as the the technical ones. And you talked about like a persona there. Is there is there somebody already in the business that potentially would have the be the perfect for that for that role?

SPEAKER_01

I mean, I'm sure there would be. There isn't a direct, I couldn't say it would be, for example, someone that's a a data analyst. It would be the things that I touch more. I mean, my the first thing, Circe, is someone's got to own it, so it can't be a side of desk project. Um, and then all the skills that we've just said, the soft skills that we've just said, if someone is existing in the business um that has those, but there's got to be investment that they own it, and that and that is what they're focused on doing. Not this is my day job, and I do this AI stuff on the side of desk because it'll it's far more likely to fail.

SPEAKER_00

Well, just a case in point, I had a chat to an um old friend of mine that's in a fairly large firm. I'll probably go as far as that to name them. Um, but um, you know, that he was in a fairly prominent AI role, but just wasn't given that um the resources or you know the the support um to do it away from business as usual that meant you know, and his he's got you know incredible domain expertise. Is he's actually gone off to set up his own business to to deliver on that because he believes it's so crucial.

SPEAKER_01

It's it yeah, it's really interesting. I had a conversation with someone um again from a very large bank who was on the procurement side of things, and they strongly felt that AI would make such a marked improvement. Uh and this bank is, as you can imagine, that as a large American bank, though they're procuring billions of dollars of value every year. And and I think he's at the same inflection point. I think he's thinking, if we can't if we can't do this, I think he he's actually considering going off going off and setting up his own business to answer this because it they're still viewing it as a bit of a risk and side of desk type work.

SPEAKER_00

Yeah, yeah, yeah. Well look, what what what are QBrick doing? And I know you talked about how you've evolved, but you're still very much looking at the talent and capability piece, aren't you?

SPEAKER_01

Well, talent and capability is still really important. Um, we've we've successfully got AI products now out there. So, and they're across various industries. One is a maintenance chatbot that allows for improvements in the utilization of aircraft. So um, we've sold that to a number of airlines, and that is a tool. That says, well, if you are repairing this, repair these elements as well around it, because you're on this certain plane, and that will increase your utilization because you don't have to go back a week later and take it all apart again. And we've got a very cool tool that shows that takes geopolitical data, FX moves, things like that to give supply chain visibility and pricing on certain ingredients. And we've sold that to a very, very large FMCG business that has pricing volatility in certain ingredients. And we've got uh quite a few things that are very specific to the insurance industry. But the thing that's really taken off recently, I suppose it touches on what we spoke about in terms of uh the persona. We've got um uh uh an AI business partner uh practice that has proved very, very popular because it's doing exactly what we said. So those people that they can't find in the business. Our individuals are trained um in AI, in this case in Copilot, to find a business value and own that. And I think the important thing there is they see it as a product, it's not a project. A lot of our clients uh have approached us with help on that because there is still an element of their mindset looking at uh projects, it's start to finish that AI products aren't like that, they are living, be breathing um uh products. Yeah, so that that that that's the thing that we're doing at the moment with our clients.

SPEAKER_00

We are um racing away here. Um and I don't want to sort of not get your view on what next really it's hard to sort of predict what what's every quarter I I I you know I run an update and and just as I do it, I'm like, my god, yeah, it's not slowing down. But you know, how should you know a P, their port codes prepare and and what might be around around the corner? Obviously, you know, it's Gen Tic and and and those sort of things. How fast do you see these sort of things arriving and and sort of just disrupting the way that we've worked for so many years?

SPEAKER_01

Well, um if you if you take Kubrick and the the pivot that we've had to make, we're only a nine-year-old organization and we've had to pivot hard into looking at products. Um within the tech within the AI technology that supports those products, it is improving nearly month to month. Some of the stuff that we couldn't do and we can do now, Agentic AI is is becoming more accessible, um, and we're seeing it um real life use cases across our clients. Um you know, I suppose to say that my view is that pri private equity really, really does have to have port codes that have AI embedded within their business because I think it's it's an existential threat if you're a if you're a company and you know, I'm not saying that AI has to run your business, but you have to utilize AI from a competitive point of view, because if you don't, I think it's an existential threat for those that are are non-users. Some may say also, um it's interesting, one of our clients has been um Fund Manager, has been saying that they are now creating their own AI use cases and agents. So they would say they are creators, not users, and their view is it's not good enough just to be users. You have to make this stuff proprietary to yourself.

SPEAKER_00

It's um I was talking to Porco the other day, and and he described it as as chaotic, um, and that it felt like uh there was a well, certainly from his perspective, a bit of panic setting in, and so layoffs taking place, but even layoffs within the AI team, it it just felt like it was just weren't sure on what what uh is the playbook really.

SPEAKER_01

Yeah, I mean we we we've seen that, and I think one of the stories that I kept on reflecting about from the AI Pathfinder event was I don't know if you remember this, but one of the private equity companies said that they had a CEO that um stood up and jokingly went to their team of 40 software engineers, went, you know, how many of you, you lot, use Chat GPT? And he became in a blind panic when more than 90% put their hand up. And his expectation back to the private equity company was I didn't think anyone was going to put their hand up. I said it as a joke, um, and then suddenly worried about the validity of the code they were writing and the quality of it. And I just thought that's really interesting because that's from the CEO, um, quite disconnected with the reality of what's happening with AI. And we're seeing that even in our larger clients. You know, there are pockets of people that are either resistant or ignorant about how to use AI. So, you know, one thing I am seeing more is you're getting far more hybrid AI type individuals that are going into the businesses, certainly the ones that I would say are leading the way in that. And I expect to see that increase. But, you know, maybe it's people just seeing back, looking back and seeing it's a cost and not getting the return immediately. But equally, I think it's when people talk about their talk about their AI strategy, it unless you're very advanced in your AI journey, I think that's a really difficult question to ask. So if you're trying to solve that, I think panic can set in because where are you going to start? Whereas if you talk about what are our two or three AI tactical wins, I think it's a lot more palatable and you're less likely to see that panic set in.

SPEAKER_00

Yeah, yeah, yeah. And and there's a big cultural piece to this, isn't there? Um, I was having a um and we won't go into this kind of worms, but um, I was talking to to uh to someone at the weekend and and um their daughters now at sort of aged 18, 19 going to work, but actually resistant to working with AI because the environmental impact, you know, there's all that and that's just one you know view of of the world. Um so yeah, there's a whole sort of cultural sort of hurdle all round to really I suppose that all inputs into the chaos potentially. But as you say, you know, it's it's going step by step, recording that, learning from it, go to the next one.

SPEAKER_01

And you know, go I suppose going back to the CEO example as well, it's who in the business is owning it. You know, we've got examples, and uh I'm not I'm certainly not talking about this in in any kind of bad way, but we've got some examples of HR owning the AI rollout. And you know, I'm guessing that this is for a relatively large finance company. I'm I'm guessing because they see um that some of the AI rollout will be copilot, and co-pilot will be an HR tool that should be used. But you know, this is this is quite a significant transformational change for an organization. You've got to be people putting people in it that understand the impact, not just kind of seeing it as a quite a ri in this case a limited a line a limited people focused tool. I'd want I'd want proper senior exec stakeholders that are owning this that as we said earlier on that are measured monthly in the board or quarterly whenever you whenever your cadence is, but measured on what is the output and what's the progress on these.

SPEAKER_00

Yep. Yep. Uh look, we could have easily gone on to speak in much more detail on any of these topics, but really enjoyed the discussion there. And uh, you know, there's a lot of takeaways for for P firms and their portcodes and what we've discussed there. They can definitely be more proactive, more deliberate, more action-oriented, and as we're seeing, there's more adoption. It's just how they do in a less chaotic way, I suppose. So thank you very much, Simon, for your for your time today. Thanks, Steve. Um, just leaves me to to plug a little session that we've got planned in September. Simon, you mentioned the uh the quarterly sessions that we run. Well, the next one's going to be a full-day event on the 16th of September 2025, a strategy day, which is specifically designed for private equity. The event offers immediate practical applications of AI throughout the deal lifecycle, supported by keynotes, case studies, workshops, ensuring you gain not just ideas but things you can action immediately. Um so I just say to check out aipathfinder.co for for more information. Hopefully, we'll see you there, Simon. And um and well, that's it. That's it for now. Thank you for listening.