John Bates Podcast Transcript

John Bates Podcast Transcript

John Bates joins host Brian Thomas on The Digital Executive Podcast.

Brian-Thomas: Welcome to Coruzant Technologies, Home of The Digital Executive podcast.  

Do you work in emerging tech? Working on something innovative? Maybe an entrepreneur? Apply to be a guest at www.coruzant.com/brand. 

Welcome to The Digital Executive. Today’s guest is Dr. John Bates. Dr. John Bates is the chief executive officer of Doxis, formerly SIR Group. 

John is an experienced CEO with over twenty-five years of industry experience, and he has collaborated three times with Doxis majority owners, The Carlyle Group. Jon is also a non-executive director of Sage Group PLC, a leader in cloud accounting and financial management software. Prior to Doxis, John was the CEO of Eggplant, a pioneer of AI-powered software test automation, and the CEO of Plat1, an IoT apps platform acquired by SAP, and founder and president of Apama, a pioneer of streaming analytics. 

He has also served as a C-level executive in several public software companies, including Software AG and Progress Software. He holds a PhD in computer science from Cambridge University.  

Well, good afternoon, John. Welcome to the show.  

John Bates: Thank you very much. Great to be here.  

Brian-Thomas: Absolutely, my friend. I appreciate it. 

And you’re hailing out of the London, England area in the UK. I’m in Kansas City, so I just really appreciate you making the time and jumping time zones and calendars, et cetera, to get here. So thank you again. And John, if you don’t mind, I’m jumping into your first question. You’ve built an impressive career leading multiple innovative software companies, from founding Apama to now serving as a CEO of Doxis. 

What key experiences shaped your journey to where you are today?  

John Bates: Well, I’m really a product guy, and so I always look at things through a product lens, I guess. I’m a former professor of computer science. I’m a PhD in computer science. And Apama, which was the first company, was when I cut the umbilical cord and went over there to… 

from the, crossed over to the dark side, if you like, and, and decided, “Okay, I’m gonna commercialize some of my research,” which was a company that used machine learning techniques to find patterns in fast-moving data and act on them in sub-microsecond latency. But so my first experience, I guess, was building everything myself from the product to the company to the go-to-market to then doing Crossing the Chasm. 

Boy, that’s a lot of effort. So I, I… We applied that technology in the in the area of algorithmic trading, so I had a lot of early experience with autonomous agents Kind of running around trading automatically and, processing millions of events a second and, and fighting against each other. 

So I, I think I’ve got a little bit of a harbinger for what’s gonna happen with the agents now that we’re rolling out in AI. But, but the– I, I learned really from that, that it’s really cool to find a really good technology and in a space that’s ripe for disruption. In that case, it was algorithmic trading. 

People were taking months to build algorithms, gave them a technology where they could plug them together in hours and p- pointing that technology in that space that’s ripe for disruption, playing into a coming revolution, and I’ve done that a couple of times since. I’ve, I’ve sold five businesses and, everything from the– my, my previous company, Eggplant, which was a really cool software test automation technology where we used… 

It was really the first AI-powered software test automation, which everyone’s doing now. But y- applying generative AI with a model to say, “Okay, let’s automate the building of test scripts based on what we can figure out about this system, and then let’s use AI-powered agents, bots, to be able to run those test scripts in real time without even access to the code.” 

So the Orion space vehicle that’s just gone around the Moon, that was– the interface of that was tested with Eggplant. So that was, that was a really cool pivoting that technology using AI into the software test automation challenge. And then Doxis is similar, which is document intelligence we call it, which is the next generation of document management, document automation and applying AI to the end-to-end document cycle. 

So that, that’s really it. It’s all about the vision and, and pl- taking cool technologies and driving them into places where disruption is about to happen.  

Brian-Thomas: That’s amazing. Thank you, and I appreciate your background. Obviously, your, your educational, your research PhD background has made a difference and contributed to your just extensive career here. 

And I liked how you – you look through this product lens as you talked about because you, you mentioned built– you built everything from the ground up. You did product, you did design, you– leadership, you did it all. And I think because of that, you built some of the most innovative companies and software and now you’re actually leveraging that power of AI in this new document automation, document management end-to-end, and I think that’s amazing, so thank you. 

And John, at Doxis, you’re focused on intelligent content automation. What problem are you trying to solve for enterprises, and why is it so critical right now?  

John Bates: The problem that we’re trying to solve really is how can you maximize your team and make them a hundred times more productive when they’re managing documents? 

And documents are really still the lifeblood of businesses. Every day, billions of documents flow into enterprises, whether it be invoices, orders, resumes, you name it. They’re, they’re, they’re driving docu… driving businesses with documents. And businesses have to store them, have to find, have to search them, have to generate new ones, have to communicate them, have to interact with them, have to, do all– collaborate with them. 

So, this is a massive problem. It’s a massive productivity problem, and there’s a whole load of legacy platforms out there that are just not bringing us into the New Era. Because what you really need to do is automate the whole document life cycle from the understanding of documents, automating that with AI, to the automation of processes around documents to, being able to store documents in a massive scale. 

I’m talking about tens of billions and being able to find things in those. How can you automatically find the gold nuggets that are the opportunities and threats in your business? And then, you know, how can you collaborate around them? How can you automate the generation of them, the communication of them, and so on? 

So, we’ve really brought together, converged a load of different technologies into one seamless platform, which is an AI-first platform. And if you just consider, for example, just take HR, which is a real big consumer of documents. Imagine you wanna hire people, you wanna understand resumes coming in, say, “Are these relevant for this job? 

This is what we’re looking for.” Match them, send them to the right people for approval, automatically triage them. Then when you wanna hire somebody, how do you automate the document processes around their hiring? How do you generate their contract? How do you send out, mass emails tailored to employees without causing massive overheads? 

And, and how can you help your team automate this? And the same is true in financial documents with the CFO department, invoices, purchase orders, purchase to pay, order to cash. Same is true in financial services organizations. Know your customer processes, detecting fraud, generation of your annual statements in insurance and, and so on and so on. 

So, documents are still the lifeblood of businesses, but they need a new era of platform, and that’s what we’re trying to do.  

Brian-Thomas: Amazing. And if you think about it, it’s just– it’s almost like we, we kinda take this for granted. There’s documents everywhere in everyday life personal or business, any industry. 

And if you think about that, it’s massive. And you talked about this problem everywhere. How do you maximize your team’s productivity when there are millions billions of various types of documents in everyday business? But you’re automating with Doxis the whole document lifestyle: receiving, sending, search, indexing, storing, et cetera. 

And you shared some examples in the HR and the finance verticals leveraging the power of Doxis, this AI-first platform, and you’re really gonna make the world a better place. So, thank you. John, with AI increasingly embedded in enterprise software, how should organizations think about balancing automation with control, governance, and business value? 

John Bates: Well, for me, it’s all about trust. Trust is, is key in, you know, in your enterprise systems. And there’s a lot of confusion right now. Everything’s about AI. You’ve seen the SaaSpocalypse in the markets and lots of fear around, “Oh, you can just do everything with, an LLM. It’s gonna replace everything”, you know. 

And maybe if you just feed your documents into an LLM that will just replace everything. Maybe it can just generate all the applications you need.” But the danger here, and I think we’re gonna see this more and more, of course, unchecked with just raw LLMs, you can get hallucinations. So you can get things that aren’t actually there in your documents. 

You can get non-determinism, so your processes will behave differently on this, on the same process will behave differently on different days, different moments, and you can generate code that can’t scale and might run into serious bottlenecks. So, I mean, for me, AI is clearly a massive game changer. 

It’s a new industrial revolution, but you’ve got to manage that trust piece. So, there’s an amazing promise out there. You can ask questions in a natural language. You can generate new applications quickly. You can use agents to replace or augment people approving steps. So but you need to add things around this. 

So, you need governance. You know, you need one security model, one authorization model. You need determinism. You need to make sure your automation behaves the same way every time with the s- with the same input. So, you need determinism in automation. You need determinism in search. So you wanna make sure that there’s no hallucinations in your search and- The, the, the way that we do that is with, a, a technique called retrieval augmented generation combined with vectorization so that you can add real data from documents back into your search. 

So, you actually, zero in on the right information rather than it, it, it cooking something up for you. You need a highly scalable system. So you’re not gonna just vibe code the, the next generation document management system. And it needs to be futu-future proof. So you need to be able to have what I call composable AI, the ability to map into different large language models even dynamically. 

So, you might send particular, analyses, requests to a different LLM depending on the specialization. And then, finally, I, I like to think of, sort of you need a checkpoint Charlie. So if you are gonna have agents to vibe code or agents to do certain things, you need to give them defined APIs to access this system so that you preserve trust, determinism, scalability, and the right routes authorization through the platform. 

So these, these are just some of the things that I think you need to think about when you’re embarking upon a next gen document plat– strategy.  

Brian-Thomas: Thank you. Lot to unpack there, but I think it’s important. The takeaway here in your answer was trust is key in these enterprise systems, especially in this world of AI with these LLMs. 

And you talked about unchecked LLMs can produce hallucinations. There could be some unchecked code creating bottlenecks, could be all kinds of problems. And I think it’s important you did highlight having that one security, one governance model. And I– g– key takeaway here is determination and automation, which you delved into there, so I appreciate that. 

And John, the last question of the day, as we look ahead to the future, how do you see AI automation and enterprise platforms evolving over the next decade? And what role will companies like Doxis play in shaping that future of intelligence business operations?  

John Bates: Yeah. Well, AI’s clearly gonna change everything, is changing everything. 

But it’s a power and like any great power, rather than just let it loose unguarded, we need to have, checks and balances around it to make sure it, it does bring in those trust. And I think that’s the kind of role that enterprise companies gonna play, enterprise platforms anyway, in their own domains. 

And I, I think there’s gonna be a, a bit of a if you like what Gartner would call a trough of disillusionment in AI gonna come up, but that doesn’t mean it’s not super important and game-changing. But I think these topics of trust, determinism, reliability, scalability are going to bite some people. 

They just haven’t thought this through. They’re under such pressure to, “Oh my goodness, we’ve got to have an AI strategy,” and, and they need to think these things through. And I think a little bit like the internet bubble where, you know, if you go back to the late ’90s, you had this, “Oh my goodness, it’s all about eyeballs and clicks,” and so on and so forth. 

And then it turned out actually the internet was a game changer, but it was actually all about value and the right business models. And so a lot of things got swept away, but the right business models did in fact, come to the the fore and, you know, whether it be, you know, your Amazons or your Microsofts or whatever the, the, the… 

still around today. And I think you’re gonna see similar. You’re gonna you’re gonna see AI find its real powerful value propositions, and I think a lot of that will be through enterprise platforms that add the guardrails around a particular domain. I think you’re gonna see domain-specific models. 

I think you’re gonna see things tailored to particular vertical or horizontal problems. I mean, all powered by, by LLMs, but with the right guardrails around it. And I think you’re gonna see also the hastened demise of a lot of legacy businesses that aren’t AI first. So, you know, that’s, that’s what I think is gonna, gonna happen at an accelerated rate over the next five years. 

Brian-Thomas: Thank you. And, and I agree with that last statement. I think legacy companies and platforms will go by the wayside if they haven’t already started working on that strategy of AI first. But jumping into a couple things here AI obviously has currently and is changing the landscape, and I think will ev- forever change the, our, our business landscape here. 

But th- and we talk about this a lot on the podcast, we need to have those checks and balances in governing these platforms. G- without guardrails, I think we could have a serious problem on our hands and that’s been a big discussion, like I said throughout the, the globe right now. But governance, scalability, reliability, those are important when you’re looking at these systems. 

So, I appreciate the insights today, John. And John, it was such a pleasure having you on today, and I look forward to speaking with you real soon.  

John Bates: Thank you, Brian.  

Brian-Thomas: Bye for now 

John Bates Podcast Transcript. Listen to the audio on the guest’s Podcast Page.