EP113: The Democratization of Automation for Mid-size companies w/ Turbotic

Hello and welcome back to the Process and Automation Podcast with the Automation Guys. Yeah, it's great to be back with another interview episode today. As you know, interview episodes are our audience favorites. And yeah, we do one today as well. And yeah, another thing everyone knows about Arno and myself, we are huge advocates for practical AI, AI which actually is used every day in companies and makes a Makes a big difference. And we are always keen to learn about how companies are using AI, learn about the latest tech vendors across the industry. And yeah, there are plenty of vendors in the market. When you look through LinkedIn, it feels like there are like hundreds of new AI and automation tools out there and sort of are created every day. Yeah, and our mission from day one of the podcast is to cut through the noise, go beyond the hype, and bring the best and practical tools and platforms to our audience. So without further ado, it is with great pleasure to welcome Theo Bergqvist. Welcome to the podcast.

Thank you so much. Thanks for having me and looking forward to this episode super much.

Yeah, very excited as well here. Yeah, so we also very keen to hear your insights today and learn a bit about Tobotic, the platform, everything you're building at the moment. And maybe to kick off, maybe you can play, yeah, share a little bit to our listeners, your background and what you do.

Yeah, sure. Thanks. No, sounds great. I'm the CEO and the founder. We started 5 years ago, and prior to that, I was heading Ericsson's data strategy. So I came from a very large enterprise automation AI program into the startup world. And prior to that, I have, you know, for 20 years building tech companies. So, so we kind of, that, that's my, that's my background. And as I said, we've been around for 5 years and been helping a lot of companies build their transformation journey through, through software mainly, you could say.

All right. Okay. What might make you think, sort of switching from enterprise to startup?

You know, the thing was that when I was at Ericsson, we saw some really big problems and not to linger too long on that, but one problem was that we could be very successful automating like processes and tasks and workflow in different functions. And by that, we could also take away FTE because therefore at Ericsson, that was the goal. But the problem was that we had to add those FTs into the IT domain, right? So we had to add people, which didn't make sense for me. So that's kind of where we started to build an orchestration platform for that. But as I said, since then we have pivoted and now we do, we released a product 6 months ago, which is more targeted to the SME markets to democratize, you could say.

All right. Yeah. Yeah. That's, that's amazing. Yeah. So definitely where it was a big enterprises, uh, yeah, things are moving very slow, isn't it? And sometimes they're very legacy tools and it requires, um, yeah, a more modern approach. So that's probably what triggered it.

Yeah, that is one problem. But I think the, you know, what is happening now, I think the last 2 or 3 years is the fact that, I mean, you guys have been probably longer in this area than I have, but one big problem I see is that to really transform your company, to really use this technology, is expensive. It costs a lot of licensing. You have to build your own capabilities. You have to, I mean, it's a new domain for many companies, which is quite expensive. So how do you, how do you kind of bypass that cost? And how do you put automation in the hand, in the hands of middle, midsize, or even large companies without a massive cost? I mean, one thing is obviously the orchestration itself to optimize the licenses, the, you know, the machines they're running on, how you schedule things, but also how do you put tools in the hands of actually people in the organization so they can do this themselves. So there are many questions I think that needs to be solved, but I think last, as I said, last 2 years, there are coming things to the market that actually enables this, which I think is super interesting.

Wow. Yeah, very nice. Okay, so then maybe, yeah, the next thing I thought would be interesting to know is where do you see the real competitive edge in the market? There's a lot of stuff happening, so where do you position yourself?

Yeah, I think, I mean, that is tricky. Just as you mentioned before, Sacha, there's so many tools in the market, there's so many AIs that can do so many different things, but we try to differentiate ourselves into one thing you could say than if we are the best or if there are other tools who can do it better. But the whole idea of democratizing automation and AI and agents, you know, whatever you want to call it, I think that is what we are good at. We can, with our new software that we released in November last year, I think we have more than 10,000, 15,000 users on the platform now. And none of them users are more traditional RPA developers. They are engineers. They like automation, but maybe they cannot code. Maybe they cannot do automation workflows in UiPath or Blue Prism or what have you. And that, I think, to close that gap, because if they can do it, next step would be that, you know, a tech savvy guy in the finance function can do it. And that is kind of our edge that we, by using Trabotic AI, you chat, you basically chat with the, with the software. We build all the integrations, all the scripts, execution code, all orchestration, and you don't have to put any line of code whatsoever to build your automation. And obviously it goes without saying that is not for the big companies, right? We all know all the compliance, security, the onboarding, but for the mid-size, I think that is really, really good. And that is how we're trying to position ourselves.

Yeah.

And I guess sort of go, going for a sort of a larger stakeholder community market where you could have these tools quite easily accessible. And, you know, I guess one thing from my side, Theo, and thanks for being on the show today, by the way. We talk a lot about kind of real-world use cases on this podcast. And, you know, our listeners are always interested in, you know, you know, what is it actually that you solve. So from your perspective, kind of looking at your customer base, you know, what are some of these kind of exciting or impactful, I guess, use cases that Tobotic is solving and delivering solutions for that you're currently working on?

That's a good question. I was in London, I think, one month back or six weeks back. And then we presented actually numbers around this because we, I think at that point we had like more than 100,000 executions the last 90 days or so into the platform. So we can, we can see what they are building and how they use the platform. Quite, quite surprisingly, I would say almost 20% of all the use cases that are built are what we classify as very hard. I mean, it's a, you know, a long, long, long processes, multi-system. We have automation built over 19 integrations, 19 different systems. If you look, so, so everything from very complex to hard to not so complex and the rest of the, The pie is pretty, pretty kind of divided among, among different levels of difficulty. But if you look at the actual use cases, what domains they are. So it's, it's a lot of marketing. We can see a lot of marketing use cases. I think almost 20%, a lot of IT, like building IT ticketing system, handling tier 1 tickets, for example. It's a lot of pure operations, a lot of, uh, Finance as well. This, you know, always the kind of the first that comes to mind with automation is invoice handling. So we see a lot of invoice handling, but not the hairier one, but more the basic one. Fetch my, you know, fetch all the invoices in my mailbox, extract all the data, put them in my simple, super, you know, mid-tier ERP, you know, giving notice to these people when it's there. Those kind of automations. We see a lot of that from these almost 100,000 executions the last 90 days. So it's a widespread, which is good for us, I think. But obviously, it's also, you know, we want, you know, to go from there to actually do the, to say, 50% is the really hairy stuff, because then you're entering also into agent domain, and we are not really there yet, I would say. But that's a, you know, a ballpark, a month back.

Okay. So just spread across kind of operational productivity type use cases for, I guess, people that have got these problems day to day that, you know, probably like you say, not technical in nature to be able to build these solutions necessarily, but they are savvy in kind of using systems. So, you know, it's by the sounds of it, your platform provides those capabilities for them then to sort of tackle these problems head-on and obviously as AI as a basis for that.

Exactly. And kind of interesting, you know, food for thought at least is that I think 80% of all our users, they have logged in with a Gmail, which means they are not enterprise, right? They are smaller, smaller companies. You know, they want to try, they want to build something. And then we see after they have built something 1, 2, 3 times, then they switch over to more the corporate account. So there's a lot of people kind of interested in this, but they couldn't do it before, but now they can.

Yeah. Yeah, I was at a conference last week. It's all about, it was all about SME sort of businesses out there, sort of high growth, fast growing SMEs. So is that sort of the market or even bigger, bigger than SMEs?

Yeah, I would say, you know, For me, I mean, it's a hypothesis. I don't know what you guys say, and I will obviously be interested to hear your view on that as well. But I'm pretty fixed on the cost for building automation, and I think it's insane that you pay $10,000, $12,000, $15,000 for a UiPath license, for example. Maybe if you're big, you know, maybe it's worth it. But if you're a mid-sized company and if you can only load it with 30% of processes, meaning utilization level is only 30%, then it's very costly. So, and SMEs, they cannot afford that. They cannot afford 20 robots going 30% allocation. They need something simpler to, to be able to automate. So for me, if I, if I look at it from, from my kind of market perspective, I think that the, the big, big boom in automation AI agents will be the money, you know, for companies working with this. Will be in the SME market, not the enterprise. But, but it's a hypothesis. I don't know.

Yeah, it's a, it's, it's, it's huge. It's a huge market. 60, uh, 60% of economy, um, sort of in general. So that's the rule of thumb is SMEs. Um, and a lot of them are between sort of 10 and 50 people. But still they have the same, they are big enough to have problems, sort of capacity problems and scaling problems. And, um, uh, yeah, and they all, they all struggle with actually translating that because they can't also afford expensive consultants going in and developers going in. Very often they don't even have their own IT department. So I think that, that could be a very powerful tool for them potentially. So they can definitely start articulating what they actually want solving and they get solved with your platform as it seems. Very cool. Okay. So, so obviously all these things, the hype topics, intelligent automation was one hype topic, agentic AI, sort of, and then it's like AI on its own. So it's very, very, um, sort of, uh, buzzword-driven currently. So every, and everything around that is moving very, very fast. Looking a bit into the future. So we all love to, to, to have that capability, but maybe you can see that, uh, you touched a little bit on that where potentially the growth is, but, uh, how do you see AI-driven businesses changing, um, and changing how they get work done going forward?

Yeah, I mean, that is, uh, that is the million-dollar question, right? If, if I, if I knew all that, I would probably put all my money on, on, on that. Bet, but, you know, I think from, from what I see and all the technology, we are pretty deep into the software development. I mean, not only our own, but to see how the models are built and, you know, all the stuff under the hood. And when you, when you look at that, I mean, I think it's inevitable that agents will, will kind of be the next big thing. But I think as well, don't forget the RPA, you know, the more traditional automation, because there's a lot of, lot of stuff to do there. And not all the processes, not all the functions need agents. And I think there is a lot of spending on agents now, but I mean, we are helping pretty, pretty, pretty big companies out there now with their transformation journey. And they also want to bet on agents, but still I haven't seen one working agent the last year. There is a lot of buzz and I don't think, don't overspend on something that is not mature yet. It's better then to actually look at what you can automate and build a more solid, solid ground. To be more effective and build capabilities. So the future, I don't know, but I do think that agents will somehow replace a lot of traditional processes even. I don't think, I usually, you know, when I talk to another client the other day, I told him that, you know, what's the big difference here between an RPA and an agent? Well, it's the orchestration because you want to know, you know, what data, you know, where does it come from? Where does it go? Agents go outside your domain. It goes outside your systems. It goes elsewhere that you don't know. What's the risk level of all these? So it's so much more complex. And I mean, it's not just let's build an RPA or let's build an agent to put it out there. Then I think that would be, you know, costly for many companies.

Yeah, I think a lot of things have not really changed that much, isn't it? So we, we are, Arno and myself, we are in the field. For a long, long time indeed. So when it comes to the next big thing coming in, then there is RPA and then there's low-code as a methodology or as a technique and as a new concept. All these things came in and the same was when I was with AI. But basically we always said you need to, you need to put the foundation in first of all, you need to start small, you find the right use cases, build a foundation in and don't go too much off-piste. And only then once you have foundations, then yeah, you can start, start growing. But nowadays I think the The hype is really getting everyone confused. So everyone is jumping on it without having actually foundations in place, guardrails in place, not even basic thinking around things in place. So that's, uh, everyone wants to, yeah, everyone thinks it's like, uh, just think it and it's done. But that's unfortunately not the situation, isn't it? Yeah.

Yeah.

It's, it's, it's always back to the roots, I think. Oh, even with AI, even if it, and, and it sounds boring for probably for some, but I think it's very fundamental.

Yeah, but I mean, yeah, sorry, Arno, go ahead.

Just to add to what Sasha said, yeah, we, from an adoption perspective, we also see some concerns about security, governance as well, even small organizations, data sharing, agentic AI. There's a little bit of a challenge actually from a change management perspective as well to introduce agentic AI and actually educate people exactly what it does. What's the domain mention? About kind of external touchpoints these agents can have. So it can be a little bit of a challenge to actually still ground yourself on those best practices where you've got a deterministic process that's well orchestrated so people can understand what comes in, what it does, what's the outputs versus something like an agent, which can feel a bit kind of black boxy and it's easy to implement, but You know, security considerations, legislative considerations, all of these things I feel is still topics that needs to be discussed. And, you know, maybe there is a little bit of a two kind of camps. One camp is let's, you know, jump into the tech, let's use it. And then there's probably another camp that's a little bit more skeptical to say, well, that's all good, but what about all of these real-world problems we still have? We still use Excel spreadsheets, we don't have kind of all of our internal systems digitized, and all of a sudden we've got this AI overlay, is that creating a better outcome for us, or is that actually creating a place that we're worse off? Agents are using the wrong type of data, so it's creating more mess quicker.

Yeah.

So yeah, it's kind of like, like I said, the points of discussion is a lot about kind of governance. The tech seems to be quite mature in the sense that there's so much out there. So if, if you make a good decision and have something that's pretty kind of mainstream, I'm looking at the Microsoft stack, for example, you know, there's this, obviously Microsoft's investing a lot in there. So if you follow that path, you're probably going in the right direction, but then you've got licensing to consider as well. You know, so, you know, there's quite a lot of, I guess, parameters in that thinking. And then you think about the change management, you know, so it is still reasonably complicated when you start to kind of scale up and then you have to think about these kind of grown-up problems, right? Rather than just having a a nice new toy and you play around with it and it's all, you know, flashing lights and wonderful. And, you know, it almost sounds too good to be true until, you know, there's a problem.

Yeah, no, totally. I totally agree with you. I don't know where you're saying that.

Yeah.

And just with that, I want to sort of just get your view, Theo, on, you know, what's your view on where people start? You know, we get this a lot from our listeners. They reach out to us and say, you know, We've got all of these buzzwords and all of these new tech. But, you know, where do we start? I'm curious to hear what's your advice to companies that want to start using automation, intelligent automation, the capabilities your product provides, AI in general. What's your view on that?

You know, I'm a simple man. I've seen a lot. I've seen a lot of different automations and I see automations that fail and I see automation programs that fail. When I see automation programs that are successful. And for me, there is only one, what do you call it, commonator, one common thing for the successful ones versus the non-successful, and that's start small. If you start small with one or two processes, you're super focused on that. You try to make it work, you know, you automate it, whatever technology you decide to, which after you've done the discovery, you start it, you build it. Because when you do that also in organization, even if you're, if you use consultant, which you probably have to do if you, if you have no clue whatsoever, even if you do that, you build capability, capabilities in-house, you understand, you know, what is this new technology? How does it apply to the processes that we are having? And then when you do that, you, you add one more and you add one more. So you, you build programming sequences. I have seen so many kind of bigger programs for the larger consulting firms. They come with a deck. You have to do 5 lighthouses, 30 automations in each, because if you don't do, you don't get any bang for the buck and you need to rip out your organization model. You need, there's so many things that, that, that the companies cannot handle, big or small, but they can always handle small iterative changes because this is as much as it, people always say it's not about technology, it's about people. I both agree and disagree because I think to some extent, it is actually about technology, but it's also about building capability and understand what this capability can do. And that lies within the people. So it's not, you know, it's not that you have to have acceptance always from the people, right? Sometimes you just need to be rude enough and say, you know, a technology can do this better than a human. Let's just do it. Let's not argue about it. But if we don't build capabilities within the companies, you cannot scale. So my advice is always start small, start small, think big, scale fast. And every time I have at least been involved with companies doing that together with us or with our software only, every time I've seen success.

Yeah.

Yeah.

I think people don't like change, right? So if you do a big program, that's big change in one go. No one likes that. Most people don't like that. So yeah, and you can only work on change in general, also on just on the people element of change if you if you do it in small steps. So if you just say, okay, now from tomorrow, we all changing, I think we lose, you lose 80, 90% of the staff, right? So, and then all the investment is basically for nothing. And maybe even worse, people think this is not my place to work anymore. And then they also leave the organization. So there's a big risk. So I don't know why still lots of organizations get it wrong. There's really a playbook out there for like 20 years, definitely across my sort of, um, out there how to do it right. But still some people think, no, no, no, I know better. All right, perfect. Theo, we have a tradition on the podcast. You probably, we talked about this briefly, but, um, you, you may be happy with this to answer a few questions to our listeners. So our listeners, um, they like, um, they like to know a bit more about, um, sort of our interview guest, not just, uh, from the technical side. Are you happy with a couple of, uh, short fire question? Yeah, perfect. Okay, so the first one we have is, who is your idol and why?

Well, I have so many idols, I should say, but one, one in particular which has followed me throughout my career, and that's Richard Branson. I think, you know, he's— for, for younger listeners, they maybe don't even know who he is, right? But it's, it's the guy who, who built the, the Virgin Empire. And I, I think starting with nothing and building such a massive, ruthless empire. I think that's, uh, I've read all his books. I'm following him on LinkedIn. I think as an entrepreneur, because I built 7 companies and I sold all of them, and one is listed at NASDAQ now. So, you know, following him, he's been, uh, you know, my, my North Star in my whole career.

Oh, wow. Yeah.

And he—

sorry.

I think we've had him before as an idol, maybe. Yeah, yeah, he, he, I think he's made an appearance.

Yeah, but he started, he started this because I think it was about a flight which was not happening, going ahead, isn't it? He was so frustrated to not get on this flight, or the flight was canceled, so he started an airline, I think, and that was the start of it.

Yeah, exactly, that was the airline, but that was later in his career. Earlier was like the Virgin Record Store, so And that is what's so crazy, going from records to airplanes and now spacecrafts and did all these adventures. Yeah, he's someone I look up to, definitely.

Yeah, amazing. And just talking about, kind of, you alluded to Richard Branson there and reading his book. A question from my side, imagine you can get all the contents and knowledge of a book, which book would you choose?

Yeah, I would probably choose, what do you call it, Encyclopaedia Britannica, right? Because it has, you know, if you know everything in there, you know, you know a little bit about everything in the whole world. So that's probably the book I would, I would—

yeah, it's a lot of knowledge.

I guess it's more than one book, but anyways, yeah, from A to Z.

If I can have that in one book, I would choose that.

Lovely. Okay, good. Thank you. So what's the best advice you have ever received?

As a business, as then it was private, you know, I guess you get from your mother and your, your, your friends. That's a tricky one, but I think one thing that I, one advice I got like 20 years ago from one of my managers, you know, I worked at Telia and I was just about to start my first startup. I think this was '99 or something. So, you know, more than 20 years ago. And she said, how do you explain that in English? Like, keep your money close to you. Don't, don't spend, because now when you're going to be an entrepreneur, even if you raise a ton, lot of money, you know, you have to keep, keep the money, what do you call it, tight or close. And that advice always comes back to me. I always very cautious about how to spend, how to spend other people's money, because that's what you do when you're an entrepreneur. And, and I think that to some extent that also goes into, you know, working with— I mean, this you guys know as well, you know, you, you do services and you help companies as well, and I do that as well. I don't want to spend a company's money unless I know that I can help them. You know, so many people do that, but for me, that's important. It's their money. They invest in me or my product or whatever, and I want them to feel, yes, it was well, well-spent money. So that, that's kind of a something I've brought with me since many years back.

Yeah. It all should be, should be the right value for it, isn't it? Yeah. So that's a bit, it's a tricky one, like keeping the money nowadays when it comes to AI.

Yeah.

Nowadays, using all these crazy LLMs, you burn tokens like crazy. No one knows really. Oh yeah, this is another 5,000 tokens. Yeah, but how do I know what 5,000 tokens means?

Yeah.

Why is it that? And then suddenly you spend millions just because you just downloaded or asked the system to export the encyclopedia or whatever it is, right? So it is crazy stuff. So yeah, you can lose a lot of money with AI if you don't know what to do. But yeah.

Just a final one from me then, Theo. If you could be an Olympic athlete, what sport would you choose?

Olympic? So I used to live in Japan, just telling you that. And I used to, or I still practice a lot the budo and Japanese swordsmanship. But unfortunately, that is not an Olympic, what do you call it, Olympic discipline, right? But if it was, I would probably, and it's called iaido, where you have live swords. Katanas. And I practiced that since, you know, since '92. So for a long, long time. But since that is not part of the Olympic disciplines, I would probably then say fencing because it's as close to, you know, to swordsmanship that you can come.

Yeah. The other one is probably a bit too dangerous.

Yeah. That's too.

Katana. Yeah. That sounds dangerous.

People go very enthusiastic.

Kendo, which is Japanese. Yeah. That I think it was some, at some point it was like a test discipline at the Olympics. So that I do as well, kendo.

That's perfect. I have one final question actually. What is your favorite AI tool at the moment? Of course yours, but next to it maybe.

It's not, it's not mine because I think we have automated as much as we can with our own tool. But it's, I mean, I'm a super sucker for Lovable. So I use it every day. I mean, I can just give you a very quick example if we have the time. Two years ago. We tried to build, or we built a new website for Tobotic. It cost me like $10K, something like that. You know, all the tools, WordPress, what have you. I had to, so first the design and then implementation and then tools. And it took me like 6 to 8 weeks, I think. This was 2 years ago. And now I can, you know, last week we built a completely new website in 2 days with a, you know, license fee for, you know, about, I don't know how much is it, $20 or $30. And that, the whole vibe. The whole vibe thing I think is made for small companies like our, yeah. So I use Lovable a lot. 11:8 I think is good as well. Yeah.

Yeah. That's very impressive. Well, are they, are they also Swedish?

Oh, Lovable is Swedish? Yes, they are Swedish. Or I think the entity is actually American, but there's a Swedish guy who has started it.

Okay. Yeah. Yeah. I thought so. Yeah. Yeah. Very good. Perfect. Yeah. Very, very good tools out there where you don't have to use a lot of money and spend a lot of money. I think you can, you can have a very, very powerful tech stack for, for under £1,000 and build a, build a startup from, from that and for the first versions. I mean, it's very impressive what you can do these days. It's a fantastic time to live in when you, yeah, don't look at all the other bad bits, but yeah. So Theo, it was a great pleasure to have you on the podcast today. So, and I'm sure our listeners would like to get in touch with you after the episode.

What would you say, fencing? Because it's as close to, you know, to swordsmanship that you can come.

Yeah, the other one is probably a bit too dangerous.

Yeah, that's true.

A katana, yeah, that sounds dangerous.

Some people go very enthusiastic.

Kendo, which is Japanese, yeah, that I think was some, at some point it was like a test discipline at the Olympics. So that I do as well, it's Kendo.

Yeah, that's perfect. I have one final question actually. We ask more and more people when I meet a lot of people as well. So what is your favorite AI tool at the moment? Of course yours, but your platform.

No, actually it's not mine because I think we have automated as much as we can with our own tool. But it's, I mean, I'm a super sucker for Lovable. Uh, so I use it every day. I mean, I can just give you a very quick example if we have the time. 2 years ago, we tried to build, or we built a new website for Tebotik. It cost me like 40k, something like that, plus, uh, plus the, you know, all the tools, WordPress, you know, what have you. I had to, to, so first the design and then implementation and then tools. And it took me like 6 to 8 weeks, I think. This was 2 years ago. And now I can, you know, last week we built a completely new website in 2 days with a, you know, a license fee for, you know, about, I don't know, how much is it, $20 or $30. And that, you know, the whole vibe, the whole vibe thing, I think, is amazing for small companies like ours.

Yeah.

So I use Lavabel a lot. 11ty, I think, is good as well.

Yeah. Yeah, that's very impressive.

Yeah, they all—

are they, are they also Swedish, right?

No, they are Swedish, or I think the entity is actually American, but there's a Swedish guy who, who started it.

Yeah, yeah, I thought so. Yeah, yeah, very good. Perfect. Yeah, very, very good tools out there. Um, you don't have to use a lot of money and spend a lot of money, then you can, you can have a very, very powerful tech stack for under £1,000 and build a startup from that for the first versions. Yeah, it's very impressive what you can do these days. It's a fantastic time to live in when you, when you don't look at all the other bad bits. But yeah, so Theo, it was a great pleasure to have you on the podcast today. So, and I'm sure our listeners would like to get in touch with you after the episode. What would be the best way to reach you?

I think LinkedIn. I always on LinkedIn, so reach out to me there. If you want to drop me an email, you can do it. It's theo@turbotik.com. You can reach me there as well.

Yeah, I put the link to the website and to the LinkedIn profile into the show notes. And I heard as well, you're very happy to give our listeners 50% off, right? When they register with the promo code TAG, I think that was. Yeah, that was, that was fantastic news we received. So we put that as well into the show notes and then hopefully a lot of the listeners are giving, giving it a try and then get, get the 50% off as well when, when it goes to the paid plans, which is fantastic.

Yeah.

So, um, yeah, with that, we are coming to the end of the episode.

Yeah.

Thank you, Theo, for joining us today. And also a big thank you to all our listeners as As always, we will be back with another episode very soon. And until then, let's automate it.

Thank you.