C2E26 - AI in consulting - exploring best practice (feat. Michael Boham)

Speaker A: From the home offices of Ash and Flow, this is Unbillable Hours, a podcast about professional services marketing. Stick around and listen to our insights, tips and best practices to improve your firm's marketing and even your career.

Speaker B: Oh, welcome everybody to this latest episode of the Unbuild Bios podcast. I think we're a bit behind with our broadcasting schedule, Ash. I'll free to admit that because I had some heavy travel season using some busy weeks as well. If I can disclose that, I won't get into it. Um, but we're back and we have actually a guest on again. Michael is here. Michael, I'll ask you to introduce yourself in a second. From a company. Well, actually you need to explain, Michael. Maybe two companies, but from a company called Kualin, which piqued our curiosity because you guys are developing custom built AI solution explicitly, I think with the objective to maybe automate parts of the permit in consulting, which we thought was interesting for a couple of reasons. One of them was that you seem to be working on the process end of consulting much more then on what Ash, you and I call the artifact and listener to the show will realize that we are often critical of, I don't know, AI slide writers or uh, AI repo generation tools because we say that's the artifact of consulting. It's not completely unimportant, but that's not very actual values the values of much more in the process.

Speaker C: No, they're just interventions. They're not really anything life changing.

Speaker B: So that's all our flattering way of saying we thought you had something more interesting going, which is why we have you on and to talk about a little bit about your perspective, right, where things are going with regards to leveraging AI and consulting delivery, where that is, where that could go, how firms get value out of it and so forth. But maybe, Michael, you say a few things about yourself first before we dive in.

Speaker D: Sure. Thank you, Flo. Ash, uh, thanks for bringing me on. I think you say interesting, but I think the interesting terminology there is very subjective based on what we're working on. So as an introduction, I'm Michael Bohap. I'm the founder of Covalent, previously Skill Size. And my career actually started as a consultant. Actually no tell. I started as an auditor, uh, in one of the big four accounting firms. And I realized very quickly after about three months that it was like the professional version of watching paint dry. So I decided to do something a little bit more interesting and going to consulting, which is where I really found my spark in sort of work at the time when I was about 24, 25 and what the. What my skill base and interest was, I was in consulting for about five to 10 years. Went from the big four to smaller ones which got acquired several times. And I started to realize I had a passion for entrepreneurship and building new innovative things. And I got tired of selling people back their own watches. So I thought about different AI was sort of emerging at the time and I sort of applying that in various different areas and facets. And that led me to skill size, which is. Which I used to work on before and that's on workforce analytics and helping organizations understand how best to channel their workforce to enact to deliver their strategy efficiently. Through that experience of working with primarily consulting firms, going back home, as it were, we emerged to realize we can do more than just look at the people. We can look at process tech and the macroeconomic factors applying to that ecosystem. So we developed covalent. Covalent is spelled Q V A L N T. But it's actually a clever play on covalent bonds in chemistry world to electrons atoms coming together to form a more powerful element or structure, if you will. And what covalent is from a business perspective is we're building AI native infrastructure for consulting. So we're latching on to the infrastructure, to their organization turning helping them do two things, as you mentioned, flow. It's optimize delivery but also generate new IP and revenue using the historically antiquated frameworks and thinking into new productizable ip. And that's what covalent does.

Speaker A: Yep.

Speaker B: And that I think so. Thank you for that. I think that's also what sparked our interest a little bit because that to me seems much more. If I may say that seems immediately. I mean we still might have questions about how different technologies. But it seems more valuable than oh, write a prompt to get a slide.

Speaker A: Right.

Speaker B: In terms of both impactfulness and also value for everybody involved. Right. The client and the firm.

Speaker A: So.

Speaker B: And since you are in those trenches or maybe since you're working on these things, can you elaborate? I mean, obviously clearly you have a point of view because that's you're building this type of product. But maybe can you comment a little bit on where do you see AI creating real structural advantage for consulting firms more broadly? Because we have, we've discussed this question on the show. There's different takes and as someone who personally has remained critical of the technology, or still is, I always make a point to not deny that absolutely there are areas where it can be very helpful or accelerate. Can accelerate.

Speaker D: Absolutely.

Speaker B: So what's your view there? Where does this stuff, for lack of a better word, actually matter and make a difference in consulting delivery?

Speaker D: In consulting delivery, I think AI creates real, as you mentioned, as you say, structural advantage when it changes, when it changes what you sell, not just how you work, if that makes any sense. So I think most firms use AI to make decks faster, but the real opportunity is turning all this fragmented information that exists publicly client data into structured, repeatable intelligence. And it's compressing the time. One of the things that we always focus on as our tagline with covalent is compressing that time to critical insight and thereby reducing decision latency. Prioritizing diagnostics that used to take, you know, diagnostics and depending on the process, could be very disruptive to bau. It could be very arduous, six to nine weeks in some cases, but compressing that critical time to kind of get in the door with what is a percent, essentially a professional trojan, of course, to two to three days or in some cases minutes. Right. And embedded is insights, embedding these insights as something you leave on the table that a client can continue to use and you get repeated revenues. So that's where I see the real value.

Speaker C: Would you also say the real value is because, you know, at the end of the day, LLMs are all pattern recognition at like ridiculous efficient speeds, right. So they will see lots of new patterns that they hadn't already observed. And even if it is observed now, you've got like data at efficiency and scale to provide them, saying that, right, this is how you've been doing it, you can do it better. Uh, you might have known this, but here's how you can do it. And I believe that's sort of aiming at as well. I mean, one of the many things.

Speaker D: Yeah, absolutely. I think the challenge LLMs and I'll have a whole different opinion about the uses of LLMs within the consulting environment. But it's about, as you mentioned, it's the realization that we can do things better. But it's not just how, what, what we're doing and how we can do what we're doing right now, but better, more efficient, but it's also what you can create a result. As a result, what are the new service lines and new products emphasize double click on that capabilities that you can leave behind with automatic, which changes your pricing model, changes your the way the thinking around revenue generation and retention and value, most importantly, and I think that's the more astute application of LLMs and AI. But then if we just really zone in on that LLM M concept. There's so many different challenges and so many different points I have about the

Speaker B: LLM which is maybe where we should be a bit clear that in the fallen solution you're developing, if I understood this correctly, you'll have to correct me. You're making use of LLMs here and there but that's not, it's not synonymous. So when you're saying you're developing AI solutions that's not the same as oh we have an LLM and it does things. But the. Actually what was the. I'm blanking on the technical term now.

Speaker D: Neuro symbolic.

Speaker B: Yes. And yet another. We had another. We discussed early on we discussed another this idea of it working in layers. What was that? It was not multi tiered, it was something else. But maybe explain that maybe quickly explain how the how does your AI solutions are maybe different. Not completely something else, but different from the how do they differ meaningfully from just your average ChatGPT subscription?

Speaker D: Sure. Uh, I phrase it in one sentence. I think we are living in a. We're drowning in information and ironically that's been propelled by AI and LLMs. But we're gasping for critical insight. So LLMs by nature of their architecture, neural networks. Right. There's a million different ways of answering the same question, which is why simple LLM wrappers. So if you go to GPT and ask it one question and go to a different chat, the question is the answer is different. Or if you go to an LLM wrapper and you click to generate the output and you click, just continue to click away, you get different results every time. Now it's feeding you with information, some of which is correct, some of which can be just fabricated but it's just there's a vast way different. There's a vast breadth of different angles it could take to give you the right answer. That is a challenge whose nations accuracy and quite frankly prompt gymnastics to get to the point that you're trying to get to. What we are bringing to the table is this new flavor of AI uh that's emerging called neuro symbolic reasoning. And what that essentially means is we're narrowing the playing field or the exploration field of AI uh compressing that neural network gateway to focus on the intent and to focus on the critical insight required. Now that's done particularly in the consulting world in adjacent to your thinking, your frameworks, your the what good looks like benchmarks to ensure that it's not just out there playing around, it's focusing that level of Intelligence, because it actually is useful stuff but within the bounds of your frameworks, within the bounds of tried and tested ideologies and theories to reduce that decision latencies. And that's what that is the core of covalent neurosymbolic AI. Do we use LLMs for other different pieces? Of course. Smoothing out the edges, create a nice fancy compelling report in and around the insights, but all in a controlled, secure manner.

Speaker C: All right, like a very strong governance framework. You have to get that kind of like discipline right, because everything needs, I mean if you're narrowing it, giving a very specific, you know, to your point, there has to be like multiple layers of governance to get it into the right place. Because not just hallucinations, you're also going to, as you rightly said, you'll get multiple responses, you get multiple outputs and that should bring its own level of interesting challenges.

Speaker D: Absolutely, that's absolutely right. Uh, that's sort of why we've taken a measured approach to that by saying hey, we have, you can come to the shop, we have our items on the shelf. Feel free to pick and use the items are tested, FDA approved and like feel free to consume them, feel free to take them for a ride. The is ready to go. Also there's a Dell approach to it. So you can come to the shop, you can buy a spare parts, you can mold it around your particular use case and get going. You own the governance for that. That's how we're doing it. We're splitting it both ways. Our versions of our tools are tried and tested with the relevant academic frameworks embedded, the checks in place and the expert validation prior to that to ensure useful accurate to from our measurement responses. However, one of the core values of our tools is you can embed your frameworks directly in our uh, horizontal architecture which would allow you to produce insights directly how you would historically with your clients, but at the scale of AI. Right. So there's, we're often doing this kind of duopolous approach.

Speaker B: So what that tells me is that the. Because my next question here in my notes would have been how do firms that get real value out of AI, uh technology, what are they doing differently? It seems like you answered some of that already by saying, um, not only do they go beyond just automating what they're already doing or trying to drive more efficiency of today's approach, they're building some new capabilities and services. But what I'm also hearing there is, they're also, there's more to applying or use the AI than swipe Credit card, get a copilot subscription and then um, shed away. But what I'm hearing here is there's quite some heavy lifting needs to be done. Whether that's your team doing part of that or I have to do that on my end. But to put those frameworks and guardrails and governance in place to really to your point, to rein in the tools and the reasoning and make sure it's both more precise and correct. Right. The hallucination thing and so forth. And maybe there's also a cost advantage because uh, yeah, it's a purposeless burning of tokens for not much outcome but it's very, very well, deliberate and very guided. So would you say this. So besides that there's some other things consultants could or should keep in mind when it comes to AI use or using of the tools.

Speaker D: Yeah, I'd say to try and summarize this and make it as concise as possible. I'd say it's a focus on this sort of tails off nicely with hopefully the narrative I'm building about what we created and why we're in market right now and still surviving. It's architecture over tools. I think the future forward looking companies, let's move out of the consulting box. They're winning by building operating systems, not prompt libraries. Right. Prompt libraries will probably be a thing of the past in the next six to 18 months because you get to a point where you don't need to write the prompts, you describe what you want to do and the AI writes a prompt for its own initiation of a task and IP over usage as I mentioned, encoding proprietary frameworks, benchmarks, data models which all these organizations have in spades historically having delivered this successfully, albeit in an antiquated environment, business environment, encoding that into your AI workflows so they can provide more sustainable uh, value and the business model shift moving away from time based billing to smart diagnostics. Key thing around subscription intelligence and embedded insights platforms. That's what we sort of provide. So we split our audience into two buckets. People just want to get going. They just want to just get going. They don't have to build a genetic code in them. I just want something that would help me optimize or augment 60% of my tasks and automate 45% of that. Let's start, let's go off the shelf. I want to, I want to really accelerate time to value and I've got the other side where we have that sort of building gene and we have this building urge and we have the agile iterative environment in our working ecosystem where we want to build our own frameworks, we want to connect these, this agent to that agent and connect from these different tools and create something more interesting. In which case we have the ability, we provide those group of clients the ability to add their own frameworks are their own benchmarks. The case studies historically tweak them in the way that it's responsive to our AI infrastructure and deliver that client insights on the basis of how the engine is set up. So that's how long winded way of answering your question.

Speaker B: But I mean coming back to this point of uh, coming back to this point, architecture before tools, I mean that's all that goes even down further. I've seen bit of this in the context of sometimes marketing transformation is a big word, but sometimes I work with consultancies where they're trying to figure out where should the marketing team bring in AI to do certain things. And not being a technical expert, I often find myself advising around the process stuff in the first place. I look at their current setup, I look at the things, the activities they are driving right now and find a lot of them um, have issues. So it's not even a question of AI or not, it's a question of what are we doing here in the first place. So it goes back to this idea of. Yeah, I like the idea of architecture. Sometimes it's even before that. Right. Strategy and what's the setup there. But I mean these are prerequisites to architecture. So again uh, included in your point. I like that a lot actually.

Speaker C: Yeah. And I really like the points and I just had a question because I'm building something separately and we can talk about offline if you like. The question here was like when you're building governance architecture in a way it's trying and automating it in a way it's to reduce the amount of oversight. But there's also a point where have so much governance architecture that it requires more oversight. And there's also the point where it's so little. So how do you balance the whole governance architecture with oversight? Because we haven't really gotten to the point where we can do away with oversight because it's still not, you know, quote unquote intelligent. It's just guided in many ways.

Speaker D: That is a rabbit hole that you will struggle to get to the depths of. With the current state of AI, I find it is easier to do to manage the AI thinking and process in parallel. So the whole idea of the human plus AI dynamic, I think that's that is the most effective use of AI. So governance is embedded in a generation. And from our perspective, governance is embedded in the creation. We will build tools. Not for a consultant to say, oh, here's Monica Valence, tools out. But here you go, this is what you need to do. We'll build tools for you to iterate with AI to get the answer right, but in a less arduous way than prompting on GPT where you say, no, that's not right. No, focus on that. No, this is a little bit more like M. A little bit more discreet, a little bit simpler and a little bit more concrete around getting closer to the relevant insight. So to answer the very question, I think it's a hard one. It's honestly, it's a hard question to answer. It's a hard place to get to given where we are with AI right now. But we can try and circumvent that by embedding as much of that as early on as possible.

Speaker A: Good.

Speaker C: No, thanks for that. Uh, just wanted to hear how it goes.

Speaker B: No, that's interesting. I mean we have alluded to a lot of this stuff. Wherever consultancies get AI wrong, perhaps also right. If I can recap what we said. I think you said if you just automate what you did yesterday, that's probably not. You're probably selling yourself short a little bit. And then we also said if you don't fix the architecture before you put the tooling in, that's probably also bad. Do you see any other mistakes in the wild? I mean, I can give you one example. I'm certainly seeing a lot, which is sometimes C firms where there is no governance or idea at all. It's sort of a bottom line like they buy 120 copilot subscriptions, dump it on their people and say go forth and figure. Right. Which I think, I mean, maybe there's something to it. If have. If it's a boutique and you have like a democratic approach to these things and you. There's even. I can even get the logic of saying, hey, the actual practitioners, the consultants that are embedded with the clients, they know how the work gets done, so maybe they have ideas. So I do understand the bottom up approach a little bit, but I also think it's probably not the way you should go.

Speaker C: Right.

Speaker B: I don't know. But any other sort of. Oh no, don't do that type of, type of things you see out in the wild, maybe?

Speaker D: Oh yeah, I can think of a few. Well, the thing is, it's hard. This is a Hard question to answer because I am not preaching from my own or singing from my own hymn sheet because I'm a massive proponent of all things AI and I'm obsessed with AI tools. For me, it's like walking into Toys R Us and I'm five years old. I want to grab everything, I want to play with everything, I want to use everything. And I immediately see value of how that can impact my work. And I'm subscribing for 10 different products to do the same thing. So that's not what you should do.

Speaker B: Which is probably also. I was going to say it's probably also another smart approach because Steve O. Will have questions why your 200 grand or your consultant suddenly also causes 50,000 and subscription fees on top of there. Yeah.

Speaker D: That margin isn't completely lost. So. So that's actually a good first point. It's tool obsession. It's buying these various different AI tools without redesigning delivery.

Speaker B: Yeah.

Speaker D: And it's putting this at the core without adapting the firm or the team or our delivery, uh, mechanism in place to support that. And you like this second point? I think it's around selling AI instead of outcomes. Clients don't care about your AI tech stack. Fine. You've got all the. Everyone's got that now. It's almost, it's completely commoditized. They care about the value points reduced, reducing risk, faster decisions and measurable performance.

Speaker B: Yeah.

Speaker D: Which aligns, as I understand a lot to what your line of work is in my case.

Speaker B: Yeah. And I think there's also. Which we can get to that is. There's also an interesting question of being smart or not smart. I will say it. So none of you have to say that you can be smart or not very smart about the impacts these tools then have on your value creation and how you handle that. You must have seen the headlines that there was a very big four company that got in the headlines quickly, half

Speaker D: a million because they, ah, asked a

Speaker B: refund from their own. They asked for an AI refund from their own auditor. And I was like, I'm not saying that's not something you should do. All I'm saying is you better only do that if you're very clear on the downstream implications of that on your own business model. If you have thought that through, fine. Go ask for the.

Speaker D: Go ask.

Speaker B: Right. If you haven't figured out how AI will impact your own sort of revenue streams and gross margins anyways. But maybe, I don't know if you have a point on where that will all take Firms and everybody else in, in terms of their margins in their business. We, we can come back to that. But before maybe for some of the listeners who are not in firms that of I think we get our fair share of very big firms so they will all have programs running and investments made. But for the people in boutiques where maybe there's not yet a fully structured plan or a bunch of resources to deploy. Any tips from your perspective on how they should think about it and where maybe they could dip in their toes in a way where they do see value relatively quickly. Not to call out copilot too often but I do hear that uh, come up a lot where people were sort of half forced by Microsoft to buy it and now they've played around with it and half a year later they say honestly we don't see what it does for us. So any better suggestions than that?

Speaker D: Well what's the question again? Is it better suggestions?

Speaker B: Sorry, where could it if you had to advise a boutique on how to get started, how to think about AI and maybe where to get started in a way that really does some, that really moves the needle for them. Do you have any tips there?

Speaker D: All right, so this is the sort of. Because on a separate capacity I do advise a few firms on their AI, uh adoption strategy and AI maximization strategy. I'd say broadly speaking three main areas. And also there's the idea of building your own proprietary technology as well or priority systems applications. I'd say start with zoom out. I'm m a product person so I'm very measured in my approach and I tend to use a very structured thinking, structured framework towards problem solving. So I'd zoom out, map your repeated insight patterns. What do you continuously consistently. How do you continuously consistently create value? Decide what should be proprietary or taken off the shelf. And that could be as simple as a four grid matrix. On one side you have impact and the other side you have potential disruption or cost. And somewhere on that grid you can start to narrow down to your field, typically the top left to get started and then use maybe general purpose AI for the more horizontal tasks. So research, drafting, summarizing. But don't use it for your intellectual edge because that's really what's separating you from Sultan B, C, D, E. You can use it.

Speaker B: You want a more purpose built tool for the good stuff, if I summarize correctly.

Speaker D: Absolutely. Purpose built tools for the horizontal stuff. Right. The stuff that sits around that allows you to get to the insight that you need. But your insight really comes from something that you're bringing to the table that's unique and that's your own benchmarks, your own flavor or thought process of the market, your intellectual capital. And that can be enhanced by AI, but it shouldn't be derived directly from AI.

Speaker B: Interesting. Yeah, good stuff. Do you have any. I mean the magic question is, uh, nobody has a crystal ball, but if you had to think about what will this do? What will the technology do to the business model of consulting? I think the question we framed in our notes was will it compress margins or not? I don't know if you want to comment on it or have a bigger picture. Um, and you mentioned you are biased towards AI and you're a huge fan. So please do take that perspective and say, hey, this is great and if it works well and if it works the way I hope it goes, where does that road lead towards? What do you think?

Speaker D: I think AI can do both. I think for the short term because of the general shock. Typically, as you see with these innovative trends, AI will absolutely compress margins for firms that are selling generic strategy, PowerPoint, heavy work on differentiated research, which is 90% of the consulting world.

Speaker B: It's a lot of it. I would agree it's a lot of it.

Speaker D: But it will definitely expand margins for firms that uh, turn knowledge into systems sell decision acceleration, which is what we're trying to do, reduce the decision latency, productize repeatable insights and operate with lower delivery cost, high intellectual leverage, which tends to be the smaller to medium sized organizations, which is why they have a little bit of an advantage in this world. I think the larger ones, the shreks of the world, they will go through a massive compression challenge because it's harder for them to change. And they're built on a lattice that is antiquated. I always say the key behavior here is agility and harnessing the concept of being iterative. And I think the larger organizations, they will probably be able to innovate at the speed of bureaucracy. Yeah. But the beauty of being agile, smaller, more smaller entity and mid size is the idea that you can probably smaller than that, iterate at the speed of thought. Which is why I personally choose to be as a smaller team than a larger one, despite having opportunities to grow.

Speaker A: But like.

Speaker B: Well, Luke and I, Luke and I. Sorry, Ash. Ash and I discussed in the previous episode. I think we made this prediction not being as deep in the technology and then also not being as deep in the market. Because I know Michael, you are working with like, we're not talking about you developing in a garage. You Guys have clients, you're working with firms, you're embedding these solutions. So absolutely. Even without that perspective, we thought that, uh, there's likely an advantage and we didn't pin it to size, but we said it's probably an advantage in deep expertise. And these pockets exist in large firms as well. They also have teams with very deep, I don't know, technical, whatever it is, understanding. And our metaphor there was to say the difference in competitive edge isn't necessary. It's not, sorry. The increment I get in competitiveness is in part a function of the tool, but that increment is sort of kept by my own knowledge. Like if you give an expert carpet in a new shiny, lightweight hammer and you give me the same hammer, the increase in value we each get from it respectively is very much limited by the fact that I don't know what I'm doing when it comes to carpentry. Right. So. So they will have a much greater advantage of that new tool as compared to the other. And I think that's where we would agree because we also then discuss the difference because it's often the boutiques that have the very deep expertise and deep specialization.

Speaker D: Absolutely.

Speaker B: Whereas the big firms tend to be more on the side of. Yeah, we also have smart people, but it's much more capacity and

Speaker D: it's not

Speaker B: just generalists, but it's more on the surface of managing and technology, management, technology knowledge in comparison to the deep, specially deeply specialized boutiques. And yeah, we also made the prediction that the AI advantage would maybe fall to them. But it's interesting to hear you highlight some of the caveats which I like. The uh, architecture first, tool second. And this idea of really rethinking what it is you sell in the first place because we didn't have the few weeks and I didn't have that, we didn't have that line of thinking. So, so it's not, it's not an automatic consequence. Right. It's not automatically, oh, I'm an expertise, I'm an expert boutique. So gimme. It'll go all well for us. You still need to sit down and think and say, hey, where can we. Beyond going just what we did yesterday, we now do a, uh, 2x the speed that's not sufficient. We need to really rethink the model even if we have the slight advantage from the get go. Would you agree with that?

Speaker D: Yeah, I do agree with that and I think just to take us a couple of steps back as well, I think the architecture comment, it's not necessarily solely on your own inherent architecture, but uh, it's what you're delivering as well. And this is something that I've observed from the most successful companies is it's deliver architecture. It's called. It's the energy grid concept. I can't remember what book I read this from, but if you're supplying electricity, you don't get more retention than that. Uh, so building infrastructure for your organizations to your clients to not be able to live without you. And that's not something that the traditional consulting model can do. The traditional consulting model worked on accountability. No one ever got fired for historically this is not the case anymore for hiring McKinsey or Goldman Sachs. But now that psychologies is shifting slightly to more value rather than the brand name. So sorry, I just had to double click on that architecture.

Speaker B: Yeah, that's an interesting add on as well, I think. Yeah, for sure. So the M. Which also means for the big guys, despite the trouble they're having, if you take their innovation to the speed of bureaucracy, also their Runway might get shorter just because the protection they have always had from the brand they have built. Right. The moats they have created for themselves might be less of a help going forward.

Speaker D: Exactly.

Speaker B: It's not going to go away overnight. Right. Time to erode. But yeah, the erosion we might say is now in full progress.

Speaker C: It's also the freedom of creativity. Like I'll give an example where I know you see coloring books and there's this guy who just uses oranges and blacks and literally creates the most amazing nightscapes. And it's just everyone's got orange and black. Everyone's. You can easily buy a coloring book, but he literally creates something that looks like it took forever to do. And so you've got like capability and expertise, but you also have to have that creativity and the ability to do that. Which is what boutiques have more freedom for when compared to large corporate structures. And I think that's.

Speaker D: And they actually, they hire more with that in mind. They hire more for diversity. I think what you find when you walk into a big four consulting firm or any other organization quite frankly, is you just see clones of each other. And I think that neurodiversity point is also not neurodiversity from a. I don't want to touch that. That realm, but it's more from a diverge. Diverge. What am I trying to get at here? It's different ways of thinking, different ways of.

Speaker B: Different ways of viewing the world.

Speaker D: Yeah, exactly.

Speaker B: Honestly, we get it. We got A bit of it. I think you even get it in the big four when they started to hire digital marketing firms and design capabilities. And so for a sudden you had tattoos, people wearing sneakers. Right. Sorry to stick to the cliches, but I think that illustrates the point. Right? You suddenly had them roaming the halls and there were collisions of culture. And that was actually, that was interesting. It was an interesting time to. To be there. But I mean, maybe because we're at time and I guess, Michael, you'll have to run. But in taking us out, I think interesting to me, and I'll reflect on that in your framing, is the fact that there still is a job for people, for smart people, for smart consultants to do, which is much more in the realm of strategy and product design, almost. So thinking about how do we deploy these tools for maximum value on both sides, client for us, that's almost like a new job. And I'm not sure that's a muscle all the firms already have. So maybe that's also a picture of the future. It's not just, oh, does the pyramid get replaced? Will AI kill consulting? No, that's. I think you could see that as a part of a new role for the humans in the mix. The humans, the people in the mix. To play is to make smart and creative choices about how to deploy the technology, which that's not about best practices and just repeating a pattern that has worked too many times. It's at the opposite end of that. Right. To your point, it's about thinking, hey, until yesterday, we used to sell this and that and have this process. Now that AI is here, what else can we do? Or how can we do that differently entirely? That might be an exciting new field and maybe a bit of a. An idea. All the. A counterpoint to all the AI going to kill consulting doom and gloom, which I'm not fully subscribing to yet, but it will for sure. I think it's clear by now, change the field or it's doing that already. Let's look at ourselves. But that could be the optimistic landscape.

Speaker D: It's shifting the landscape. I think AI has a greater chance, as it's demonstrating right now, of killing the software space than it does consulting. Everyone's. Everyone there's. As long as there's change, consulting is safe and change isn't going anywhere anytime soon. It's just the model and manner which we deliver change. That's going to change, that's going to adapt.

Speaker B: All right, yeah, fair enough. But I guess we all have to scoff and run two different things. So thanks very much Mike for hopping on.

Speaker D: Yeah thanks for joining very much.

Speaker B: Interested in learning more about you exploring the specific technology. We already touched upon it briefly of behind quality and what that does. Where would you point them to? What are the links we should put in the show notes I guess.

Speaker D: Sure my LinkedIn profile works best. I'd like to say TikTok or Instagram but I'm not quite that cool. LinkedIn works. Go to website covalent.com qvirlnt.com or flagship skillsize IO.

Speaker B: Brilliant. And with that I think Ash, if you don't have any questions left I'll just stop the recording and wish everybody a happy weekend and your start to their respective weeks. Anything else?

Speaker D: No, not for me. Thank you so much. Cheers guys.

Speaker C: Thanks for joining us.

Speaker D: It's been fun.

Speaker B: Thanks for hopping on. Thanks for listening everybody and we'll speak to all of you again next week.

Speaker A: Thanks for listening to Unbelievable Outs. If you want more, tune in next week, you know where to find us.