Building the AI-Native Collaboration Layer with Clyde AI

GM, GM. Welcome to what's going to be another fantastic episode of DevNTell. So if you didn't know, DevNTell is a 30-minute podcast held every week allowing founders, hackers, and anyone in between the opportunity to come on the show and showcase what they built. And today, I'm ecstatic to welcome Tim Lidman, who is the co-founder and CEO of Clyde AI. So if you're not familiar, Clyde AI is the AI-native collaboration platform for humans and AI to work together on real problems and drive real outcomes. So if you stick around for today's episode, you'll get to meet Tim, get an introduction of Clyde AI, and learn how you can start using them today. All right, let's get into it.

Good morning. Welcome to the show, Tim. I'm ecstatic to have you on today, man.

Yeah, thanks so much for having me.

My pleasure, my pleasure. Happy Friday. Yeah, I know you got a lot of great content for our audience here today around Clyde AI. You know, AI is all the rage, so I'm sure our audience is chomping at the bit to learn all about the platform. But before that, would you just like to give a introduction about yourself?

Yeah, no, absolutely, absolutely. So yeah, good to meet you all. I guess start out with that I found my way to being a tech CEO in kind of an unconventional way. So I eventually or initially wanted to be a rock star. I had hair down to here, was kind of an anarchist, a heavy metal drummer, moved to London, and then kind of through a series of steps accidentally found myself in tech. Worked for a number of big tech companies early on inadvertently, because they kept kind of getting acquired. So I was at Webex, got bought by Cisco, didn't want to work for big Cisco, so then went and joined SuccessFactors, they then got bought by SAP, to the point where I eventually wanted to do a startup for real.

So ended up running ThinkTank, which was sort of a labor of love for eight years, which eventually got bought by Accenture in 2021. And ThinkTank is quite relevant actually to the business I've now built, which is Clyde. You know, ThinkTank was a collaboration tool that was very much designed for client engagement in consulting firms. So kind of instead of showing up with a myriad of tools like PowerPoint, Excel, Miro, Mural, whiteboards, you showed up with ThinkTank, which in a very structured way allowed you to kind of co-create outcomes with your client, had all the collaboration features you would need to get to aligned decisions and outcomes. So our largest clients were the largest consulting firms in the world, of which Accenture was one and ended up acquiring us. And then at Accenture, I spent four years operating as a partner and kind of got to see the real inner workings of the consulting industry and was, you know, pretty surprised that a lot of work was being done very, very manually and using some very analog tools. And so then when AI hit in 2023, you know, 2024, I had this hypothesis of, you know, could you train an AI on, you know, kind of the workflow of consulting to take away a lot of the really heavy lifting that goes into, you know, designing client engagements, facilitating them, you know, the poor people that are up till 3:00 in the morning putting PowerPoint decks together for clients, things like that. And so ended up leaving Accenture, brought in a dream team that I've worked with for 12 years, and then we set out to build Clyde. Raised a bit of money, we were able to get an MVP out that was very well received. And then yeah, launched the commercial version in April and excited to be rolling this out into some key strategic consulting firms.

Amazing, amazing. Yeah, I mean, there's a lot to unpack there. I want to kind of rewind it back to your story around how you started as a heavy metal drummer, because that's quite unorthodox, but everyone has their own origin story, right? What kind of inspired you to take that leap from musician to somebody in tech and eventually a founder?

Yeah, so it was, I mean, somewhat accidental in the beginning, because when I was living in London, you know, just moved there from Sweden, which is where I'm originally from, I had a problem, which is that my rent was higher than my pub job was paying. And so I was really struggling to make it financially. And then through a couple of people I met, it was suggested that I might be good at selling technology and selling software. And so then I got a job trying to convince—this is how dated I am—trying to convince people to meet on audio conferences instead of in person, right? So it's kind of the birth of the collaboration industry. And I kind of fell in love with it and just loved working with clients to help them change the way they were working and ended up doing quite well pretty early.

And then as the collaboration industry evolved, right, and so as broadband became a thing allowing for web conferencing and video conferencing, I sort of started evolving together with that industry, and then saw that people were just solving more and more complex problems and learning how to work together in these new ways. And then from there, as the industry evolved, I got into enterprise social networking, right? This is like when Yammer was bought for a billion dollars by Microsoft. I, you know, this is when I joined SuccessFactors. And then I started seeing how the industry was evolving from just connecting people to now also trying to use collaboration to solve specific problems. And then from there, what I realized, and this is kind of where ThinkTank really came into the picture, is I realized that the industry was getting to a place where we were going to fundamentally be able to start connecting problems, people, actually getting to the heart of why people engage and collaborate, right, and the humanity that's involved in getting people to an aligned decision and being able to all move forward towards a shared goal. And I just found myself increasingly being really passionate about that.

And then what I realized was that, you know, when you're passionate about something to that degree, it's very, very hard to effect the change fast enough in like a corporation, right? Like it's just moving the ship, steering the ship, it just takes too long. And so then I had several realizations and kind of inflection points where I was just like, "Okay, I think I need to be an entrepreneur. I think there's a way that you can get things done in this kind of quick, dirty, guerrilla way that you just can't do if you're an employee." And then, you know, learned that a little bit at ThinkTank increasingly. And then post-acquisition of Accenture, you know, when I saw that not only am I passionate about how people collaborate, but now there's like a $1 trillion industry that's about to get disrupted off the back of it, then I thought, "Okay, this is like everything I've done has converged to this moment now where the timing is right." I don't think I'll ever have better market timing in my career ever than what I have now. And so the stars kind of aligned, and then I had to just take the big leap, and yeah, that's kind of where I'm at.

Amazing. Yeah, yeah, absolutely. And to your point, there's no bigger opportunity than now. AI kind of lives in its own little bubble, I would say, on X and to a degree on Reddit. There's people like you and I who are kind of like on the frontier, like, "Oh, we realize how great it is." But like to the average person, it hasn't really totally clicked yet. Like people are still kind of just starting to understand like, "Oh, there's more to AI than just being able to read my emails and fix my grammar and whatnot," right? So yes, there is indeed an entire sleuth of opportunity here.

And kind of going through all these experiences and you went to ThinkTank, got acquired by Accenture like you mentioned, so you lived that big game consulting life, and you kind of saw and lived through the consultant's lens day-to-day. And I'm sure you noticed a lot of repetitive, monotonous things that a lot of consultants have to do throughout the day. So just for our audience who might not have a consulting background or work for a consulting firm, what are some of those pains and things that you experienced and kind of led you to starting Clyde, sorry?

Yeah, no, absolutely, absolutely. I think so it's kind of interesting. You know, there's a saying—in Swedish it's a bit of a different saying, but I know there's an American equivalent—which is that like the shoemaker's kids has ugly shoes. I don't know if there's like an equivalent, right? Where basically the people in the consulting industry that are helping corporates and organizations solve like the hardest problems, right, they don't get the time to solve their own problems, right? Like they just don't get the time to internally take care of themselves. And so one of the big, going to the specific challenges, is, you know, your average consultant has historically been just very, very overworked, right, in that they have to do kind of everything, right?

They have to do very manual tasks such as, you know, wrangling and collating lots of data input, right? So getting data from, you know, stakeholder interviews, workshops, disparate like emails and sticky notes you can't read and understand. And so there's this challenge where, you know, the poor analysts that are kind of out of college, they spend a lot of their first couple of years in consulting just doing a lot of grunt work, right, that's just really taxing, right? And obviously AI has a huge opportunity to alleviate a lot of that just kind of non-value add, human-heavy, expensive work. So I think that's one sort of big problem. The other big challenge is that, you know, consultants are constantly having to kind of reinvent the wheel every time they do like a customer project. And so a lot of consultants spend their time scouring through like old deliverables, you know, just trying to find where the right information lives, and a lot of just kind of detective work and scrambling together whatever thought leadership they need to sort of put the right proposal together, and just burn a lot of time trying to, yeah, align with tons of people and reinvent the wheel, reinvent the wheel, reinvent the wheel.

And then I think the last one is—and this is maybe more personal to a consultant—is that they want to work on interesting problems, right? They want to intellectually stimulate themselves. They want to help customers. They want to deliver lots of value. But the monotonous work gets in the way of that. And so they only get to spend maybe 5 or 10% of their time actually doing that. And so I think that with AI, you can look at things through two lenses. You can either look at the promise of AI and say that, "Oh wow, imagine what all this talent can do and all the value-add work they can focus on if this is off the table, right, and look at how much they can grow and enjoy their jobs and love what they do." And that's kind of what I support, right, and what we've built in Clyde is very supportive of facilitating what humans should be doing very well and then automating what they shouldn't be focused on. And that's kind of my positive, you know, the sort of bull case for AI in consulting.

The evil one...

So your background there, the Sauron version, is basically that, "Oh, we can just cut all of these analysts out and just change the talent pyramid so that, you know, we can get a ton more profits, we can deliver faster, better, cheaper, and we actually don't even need, you know, any of these resources anymore," which I'm, you know, non-supportive of because I don't think that'll drive innovation or growth. I think that'll just constrict the industry and sort of just make everything about, you know, cost and efficiency and take out. So yeah, that's a bit of a background.

Isn't that called Skynet?

Isn't that called Skynet?

That's called Skynet! That's called Skynet, indeed.

So yeah, we'll see how that plays out. Hopefully not like the movie. But yeah, kind of bringing it back, so you've kind of gone over all these different types of things that kind of make a consultant's life a bit of a pain, those long nights having to work, having to use that midnight oil so to speak. So that kind of, as far as I understand, kind of played into a lot of the core components that make up Clyde to basically alleviate a lot of those monotonous tasks and kind of create a more collaborative environment. Because I know I've taken a look into Clyde itself and it seems like you got this whole roster of AIs that you can collaborate with. But I'll let you explain and give us the pitch, so to speak, of Clyde.

Mm-hmm. Yeah, so Clyde, at its core, what Clyde's trying to do is, kind of like I was saying, what Clyde's meant to do is help bring out the best in what humans are supposed to be best at, right, and then automate what AI is frankly better at doing. And so the way that translates into the actual product is, you know, as you go through a consulting project, right, there's always different kind of phases depending on what the project is. And in some of those phases, you have to do things like, you know, a lot of data gathering and data input, and you have to kind of, yeah, synthesize very large data sets, right, and come up with, you know, hypotheses and deliverables and things like that.

But then interspersed in doing that, you also have to engage with your client, right? And you actually have to like, even if an AI can generate something in like 30 seconds, right, which is like table stakes now across all of consulting, you still have to engage with your client, get them to first understand what it is that you're working towards, and then they have to buy into it, right? And then they have to make the, you know, decisions around like, "Yes, we support this hypothesis," or, "No, we don't." And then they have to buy into actually then implementing that decision and bringing everyone along kind of the change journey that comes with that decision.

And so what Clyde does is sort of when you first start a Clyde session, you know, the first thing it'll ask you is, you know, "What are you trying to achieve with your client?" Right? Like, "So what's the problem you're trying to solve?" And you give it the initial problem, which could be, "My client needs to improve their sales process," or, "My client needs to increase revenues," whatever it might be. And the first thing Clyde will do is kind of work with you to make sure that the problem statement is very well-defined. And so, you know, if you think about some of the other LLMs, they kind of rely on the user to provide really good context, right? And it's almost a skill, right? Like to be a good context engineer and, you know, make sure you prompt correctly. Well, Clyde actually tries to take that heavy lifting away from you and actually helps to extract like your intent and your context.

And so that's kind of the first little part. Then the second thing, once Clyde has kind of figured out what you're trying to achieve, Clyde will then also try to figure out, "Okay, who needs to be involved to solve this problem?" Right? And that's a combination of AI advisors, so we have this library of AI advisors that have a raft of different expertise and personas. Also consulting firms can create their own AI advisors. So if a consulting firm has a specific expertise or a specific way of working, they can actually customize an AI advisor to be, yeah, specific to that consulting firm.

And so then what Clyde will do is Clyde will scan, you know, the out-of-the-box advisors plus any custom ones that have been built, and will recommend, "Okay, to solve this problem, let's bring in these four AI advisors." And then we'll add them to kind of the workspace. Then it'll say, "Okay, what human beings do we need?" Right? So, you know, there's two main reasons that we need people in the process: one is that they have some tribal knowledge, right, some institutional knowledge that AI just doesn't know about, right? And so then we need to bring in those people into the session to extract that knowledge that only they have, right? The other thing you need them for is to, yeah, buy into the actual decision.

And so then Clyde will work with you to figure out who do we need information from, who needs to make decisions, and we'll then invite those people into the workspace. And this is where it becomes collaborative, because they get a link, they join, and then they're in the workspace together with you. And then the last thing Clyde will do before it kind of gets into the execution of it, is Clyde will then say, "Okay, what are the series of activities that we need to go through to achieve the outcome?" And this is where those activities are split between some things that are very AI-driven, right, where AI is synthesizing, the AI advisors are coming with their perspectives and ideas. But then there's also activities inside of Clyde that then have the UX that allow us wonderful humans to type in our own ideas, theme them, organize them, prioritize them, right?

And so the Clyde has the functionality to say, "Okay, we've gathered all this information. Now let's do, you know, a vote together on which one we find is most important." AI can then go and synthesize that, come back, and we'll do a trade-off analysis, right, to then come with a recommended decision. And so Clyde is constantly facilitating and guiding both the AI and the humans through this kind of flow where at the end of it, what you get is you get an output that is then very specific to the deliverable of the consulting project. And then the last thing I'll kind of mention is on top of all of this, what a consulting firm can do is they can customize Clyde by implementing what we call guardrails where they can say, "Okay, this is as a firm how we show up to our client, right? These are the typical things we do, these are some of the outputs that we need to use."

And then you can basically enforce that so that Clyde will help you drive consistency across all of your consultants. Because at the firm level, one of the biggest problems with AI right now is that everyone has their own little Frankenstein toolkit of AI, and so they're showing up to the clients very differently, and there's no consistency, and the quality levels are kind of all over the place. And so there's really no ability to drive a high-quality consistent kind of experience, which is also introducing a ton of risk into these like huge $100 million type programs, right? Because how do you stay on top of mitigating all this risk? And so by having these guardrails and what we call team memory, Clyde actually helps solve that problem as well. So yeah, that's about as quickly as I could explain that. And yeah, to you non-consultants out there, hopefully that's making some sense.

Yeah, yeah, and that's quite a feature-packed platform. And yeah, that angle of being able to bring in other people, other relevant subject matter experts into an ongoing conversation with Clyde and the sub-agents is pretty neat. I think OpenAI had something like that a year or so ago. It didn't really catch on. I think it's because it didn't really have a niche use case around it. But something like this, or maybe in travel, like having a bunch of people figuring out a trip with AI, it totally makes sense around that. Coming back to the guardrails that you mentioned, those guardrails, those are Clyde-implemented guardrails, or can those be tuned for the end client? Say, I'm running a company, I have a specific set of rules that I don't want this AI to kind of go outside of, can I specify those as a user of the platform?

Yes. So we have some platform-level guardrails that are designed to really follow the general workflow of a consultant, right? So it's kind of very much built for how consultants work. But then on top of that, a firm almost always has like their secret sauce, right, of how they want to go through it. So yes, each firm can customize it to how they want to show up, either as a way of differentiating themselves, right—so this is, "My Clyde is better than your Clyde because we have all this secret sauce"—or to actually try to mitigate risk. Because AI introduces a lot of risk, and so to be able to put guardrails around it to say, "As a firm, we do this, we don't do this, and we need to do that," it's a way for the leadership of consulting firms to actually let their people loose in a way that doesn't just cause, you know, kind of random mayhem.

Absolutely. And I guess for our audience watching today who want to give Clyde a spin, how can they get started? How can they get involved?

Yeah. So if you're just an individual, consultant or not, you can go to meetclyde.com and you can sign up. There's a free account, and then you can upgrade to individual, you know, kind of paid versions. If you're a consulting firm, then, you know, it's best to get a demo from the team where we can walk through specifically how this applies to a consulting firm, and then we have pilot options of how to then look at this at kind of the firm level. So those are the two ways.

Cool. And yes, for our audience, all those details will be in the description of the podcast below. So if that is something that interests you, definitely give that a check out. And as we're kind of coming to time here, Tim, as we learned from your extensive background and founder journey, I'm sure you've learned a lot of lessons along the way. And for our audience watching today who might be just starting out their founder journey or on the cusp of starting a company for the first time, do you have any words of advice for them? Maybe it's some words of advice you'd give your younger self?

Yes, yeah, there's a few. One is a dear friend and mentor of mine always told me to, "Don't ride the roller coaster," is kind of what he said, which is that there are so many highs and lows with running a company. And of course, you want to enjoy the great moments and mitigate the bad ones. But if you're on that roller coaster all the time, like you will emotionally burn out. So you need to find a way to have some distance between those highs and lows and then being able to just operationally react to it. So I think that would be one: don't ride the roller coaster. And then the other one is—and this is going to sound like just such a cliché—but don't give up when you think you've reached like the limit or the max. Because the whole quote, I think it was Mandela's originally, the whole, "It always seems impossible until it's done," right, is so true. Like there are moments where I have just seen no way out, right, and like everything's awful. And then you just keep going, and then suddenly you find the little escape hatch, and then you move on to the next thing. And I truly believe that most founders who fail, they just gave up a little too early. And so just stick with it.

Never give up. Never give up. Yeah, there's always a way to do what you want to do. It's just you gotta go that extra mile so to speak. But yeah, Tim, it was a pleasure having you on today. Awesome learning about Clyde, as I'm sure our audience did as well. And just want to wish you the best of luck continuing to build out the platform. And yeah, only see bright things ahead.

Great. Thanks so much for having me on.

My pleasure, my pleasure. And with that, I just want to wish everybody a happy Friday, happy weekend wherever you may be. And you'll definitely want to stick around on X and YouTube because we will have another episode of DevNTell for you in a few hours' time. So a rare Friday doubleheader. But with that, Tim, thank you again so much. Appreciate it. And yeah, we will catch you back here for the next episode of DevNTell. Till then.

Awesome. Cheers.

Cheers. Have a good one, folks.