This is an AI Transcription. It’s pretty good, but please forgive any errors.
[00:00:00] Jonathan: Welcome to the Difference Engine, the show for tech founders, investors, and innovators.
If you go to today's guest's X account, it says, in part, moved from the Himalayas to London, learnt a bit about endpoint management and software, trying to give back through my foundation. To paraphrase Winston Churchill's comment on Clement Atlee, Samir Khairi is a modest man with a lot not to be modest about.
In fact, Samir, is a globally recognized expert in managing and securing Windows environments.
[00:00:37] Paul: Yeah, and he founded his company, 1E, in 1997 in a flat in the London suburb of Ealing, with the goal of reducing the cost of managing Windows PCs and servers through automation. He's since led 1E through incredible growth, selling most of his business in 2021 to the enormous Carlyle Group.
After having built it organically for 20 years, it now boasts offices in the U S Ireland, Australia, and India, and serves 1, 700 organizations. 42 countries with 26 million licenses deployed.
[00:01:10] Jonathan: The company now ranks in the Sunday times international track 100 league of the top private companies, and has been marked as one of the top 20 companies for CIOs to watch by CIO magazine.
Customers include some of the world's largest companies and the military of several countries. Still based in London, Samir is a passionate believer in philanthropy and charitable giving. Samir is also a judge of EY entrepreneur of the year. and a member of the Forbes Technology Council. Samir Khare, welcome to The Difference Engine.
[00:01:44] Paul: Hey Samir, how are you?
[00:01:45] Sumir: I'm very well.
[00:01:46] Paul: Delighted to have you. You and I have worked together in the past, and one of the things I happen to know is that you care passionately about, uh, green. And in the age of ESG, every firm thinks they've got to make announcements about why they are mitigating carbon and why they're doing the right thing for the planet.
But in particular, with the cost of IT operations these days and the demands that they put on licensing costs and labor, energy is rising to the top. And now the spotlight's more than ever on cloud compute and AI. So I just wondered when you envisioned efficient computing, How did you approach, you know, mitigating the costs to the planet of the IT you're involved with?
[00:02:27] Sumir: So actually, I've got a great story on this because, um, we wrote, I think one of the first green IT pieces of software called Nightwatchman. Um, this was, uh, about 20 some years ago. And, um, what it did was switch off computers. In those days, people used to have desktop PCs, which used, uh, Lots and lots of power and every desktop had one of these and people used to leave them on at night.
And so we started switching them off using the software and we could switch them on using software as well over the network and no one wanted to buy it. So, um, eventually we came up with this idea, which was, let's just go directly to the public. And so we did a study in the States, basically said in corporate America, you could save 2.
8 billion every single year just by switching off just corporate America's computers. And suddenly the phone started ringing. Right, because this was, this was picked up by Bloomberg and Good Morning America and pretty much everyone. And it was really quite interesting to me when I tried to give it away and I really did try and give away this technology initially for free.
No one was interested. Now we were charging some money for it and, and, and we were selling quite a lot of it. And actually people were very interested in, in the reporting side of it. So I think one of the things I learned was that IT have their own function. IT's function is not to save energy. Right. And frankly, You cannot expect IT to do those kinds of things, but most organizations today have reporting function directly to the board.
And actually, as long as that function's working, they tend to be people who are highly motivated. They tend to be people who want to do the right thing because they've actually, you know, these people have, have studied, you know, they've got great degrees. They're reporting directly to the, to the board. So, you know, these are bright people.
Right. And, and generally speaking, if you can get to them, um, and you look at what they are doing, you're trying to do the right things. So I think there is some greenwash. Um, but actually I think most of the organizations that we are, you know, that you talk to, you will have these people who really do have a mission to save the planet or try and do their best in, in not damaging the planet as much as we, possibly are right now.
And, uh, I'm, I feel very positive and optimistic about these people. These are the people who are going to make a big difference to our corporates. We saved about 4 billion of energy, and that's just by switching off computers.
[00:04:47] Paul: One of the things that, that in your little story about when you started charging, people started to come that you often hear.
I hear a lot on, uh, advice for SAS companies. The VCs will often say just double the price. You know, it's less hassle, fewer deals. And, uh, Maybe more money and you give an intrinsic value of it. It's interesting that the energy efficiency part of it, which may have appealed to you as an engineer, only became appealing when you converted it to two dollars, is that your experience with, with money things?
[00:05:16] Sumir: Absolutely. You know, one of the problems with being an inventor. And, and, and being an innovator and we had like 40 some patents in my company, you want everyone to use your tech. So you don't, you sort of keep thinking if I give it away, maybe they'll just use it, right? Because that's really your purpose.
But, and, and then, and people don't use it because they don't value it. And so, um, pricing is one of those things that I would say that most entrepreneurs really should try and spend a bit of time on, right? Because it's a really hard one. Yeah. And it took me a very long time to try and understand how pricing works.
[00:05:49] Jonathan: And you think this is particularly, particularly the case when something has an environmental benefit, working out what the pricing should be, can be particularly difficult there.
[00:05:58] Sumir: I think it's, it's to do with pretty much anything in. with any product really, right? But I think in the environmental space, the challenge is there is a lot of greenwash, right?
There is a lot of, of claims, et cetera, being made. And you know, how many sort of trucks do you see going by you these days that say they are zero emissions trucks? Now I know that's not right. They know that's not right, but they have it plastered all over themselves. I, I find the problem is trust. And I, I think how do you deal with bringing trust back in, into most of our conversations, into our relationships, et cetera.
I think we're sort of really lacking that in the AI space, in a lot of spaces. But the green, green, um, certainly around greenwashing, I mean, there is a complete lack of trust right now. If you are going to save energy, or you're going to try and do the right things, Then be completely and totally transparent about the benefits and the drawbacks, et cetera.
I think if you are that, then you find that people actually, you know, have trust. They're willing to spend. I mean, at one point we were selling You know, we were doing individual deals that were four or 5 million, right? Yeah. With some of the larger American companies has just on saving energy, but a bit of software that just saved them energy.
And they were willing to do that because we could prove exactly how much energy we were going to save before we even saved it. And then we proved it after the fact they could measure it. It was a remarkable thing that Dell did. They actually didn't tell us till they did it. But they wanted to sort of report back to the Congress about what they were doing.
So they started measuring energy in some buildings. And our savings came up as exactly the savings that the building was showing when our energy savings were turning on.
[00:07:46] Paul: It's hilarious, because that entire type of software that you're talking about has now its own category. It's called ITSM, IT Spend Management.
And there are big players in there, you know, VMware being one of them, now part of Broadcom. Of course, and everybody has an offering. We worked, we had the pleasure of working with Aptio for a while. Their version of ITSM was TBM, technology business management. But that small little innovation that you guys thought of and sold so successfully is in fact its own category.
[00:08:20] Jonathan: Next up, let's have a look at the CrowdStrike debacle. As most listeners will probably be painfully aware, recently we experienced what has been described as And I use inverted commas here, the largest outage in the history of information technology and historic in scale. Now it occurred because US cybersecurity company CrowdStrike distributed a faulty update to its Falcon sensor security software that caused widespread problems with Microsoft Windows computing.
running the software. As a result, roughly 8. 5 million systems crashed and were able to properly restart with the worldwide financial damage estimated to be in excess of 10 billion U. S. dollars. So Samir, not a good day at the CrowdStrike or Microsoft office. And it strikes me that this debacler seems to be emblematic.
What you've been right at the heart of ensuring doesn't happen.
[00:09:25] Sumir: Let me just start here. I'm shocked that CrowdStrike are still a business. Now, actually they should never have happened. We trust security firms with root access with basically the highest level access to our, our computing systems. Because we believe that they are going to protect them, not take them out.
And that means that there's a high degree of responsibility that they have in not causing blue screens of death. The reason that's close to my heart is because my last company was called 1E.
[00:09:54] Paul: 1E is the error message, right?
[00:09:56] Sumir: That's right. So when, when you get that blue screen of death, you get a zero X and then long number and 1E.
Um, and that, that basically meant someone, some third party, i. e. not Microsoft, had actually made a, a bit of a big blunder, as in not tested the software, like CrowdStrike. Um, and so, It should not happen. It absolutely should not happen. And it did.
[00:10:19] Jonathan: Am I right in thinking it's incredible that Microsoft relied on a third party for a totally mission critical aspect of delivering their service?
[00:10:30] Sumir: So I think Microsoft's completely abrogated its responsibility here. Um, you know, if you look at Apple and Google and their stores and the way that they manage deployments to their operating systems, they basically have a certification process in the Microsoft world. There is no process, right? Even now, there is no certification process.
Even people who have route access like CrowdStrike can just distribute a piece of software that crashes. Eight million computers and Microsoft say I wasn't awful, right? I mean, how does that work? It's really is time that Microsoft sort of started taking some responsibility. And if you're going to have systems that can crash or software that can crash all Windows PCs or many Windows PCs, then they must go through some level of certification.
And if they don't. Then actually to be quite honest, the market force is not quite working as far as I'm concerned, right? Yeah. CrowdStrike really ought to be punished for this. And yet actually right now there are big buy symbol, right. For, for a lot of bankers. I suspect actually it's not going to quite pan out in their favor.
I suspect what's going to happen is a lot of the renewals are not going to happen. And those people who have renewed till now with CrowdStrike are probably going to go over to Microsoft. So Microsoft's not particularly unhappy about this. But this whole situation is terrible. You know, how many people missed their flights that day?
Now, how much other sort of aggro was caused, and this could have been entirely prevented by just another level of testing.
[00:11:59] Paul: Yeah, I mean, I've got a, um, sort of maybe I'm not fighting the other corner, but I have to point out, um, I've got personal friends, uh, quite high up in CrowdStrike on the marketing side of things, and I'm sure this wasn't anything they planned, but also.
And maybe moving away from the CrowdStrike piece, but, you know, specifically, there is a movement in cybersecurity, obviously, you played in this space, to talk about platforms. And the idea of a platform is that it solves many, many issues, does many, many things, and you buy from one vendor. Now, vendors like that because they take a biggest, bigger slice of the pie.
They get a big, bigger wallet share, as they say, by claiming they're a platform. The truth might be somewhat different to that. There may be good bits of your Platform and bad bits of your platform. But what they're doing is making it easier to buy. And certainly CrowdStrike and others have been claiming for a while that they are the one security platform.
To rule them all and I just wonder what your thoughts are as somebody who's built really deep technology that does specific things About the claims of platforms when it comes to dominating a category.
[00:13:06] Sumir: It's a great question There used to be a time when we used to build products and then pretty much every marketing Product marketing person sort of turned around and said, ah, you can't call a product anymore.
You know, this is obviously a platform, right? and um And I think CrowdStrike basically, you know, decided that they had a platform as well. They invited people to build some scripts, et cetera, on them. And pretty much everyone's doing it, right? If you can run a script, then you basically say to people, ah, we are a platform.
Well, you're not. The fact is CrowdStrike does one thing quite well. Till till a certain point and they don't really do a lot else to try and own a category by calling yourself a platform. They're two completely different things. And if you look at the Gardner sort of model of the category, which comes much later than you guys, I mean, obviously you're the category experts, but you know, where Gardner is, you know, they're saying is enough money being spent for us to call this a category, but way before that, right.
There are people who are making. Decisions to actually create something new, to actually solve a problem that is categorical. Thinking about categories actually changed the way I was thinking about software really. And, and, and also I think the whole platform thing is nice. But how many people really achieve a genuine platform where other people come and sit on your platform virtually?
No,
[00:14:27] Paul: it is completely different to solving a deeply held problem with a really well crafted solution So yeah, I totally agree and I think that the problem I have with this platform Um platformization, let's say of tech is that you know, I don't know if you made many. Um, Acquisitions, but people acquire and then chuck it onto the platform like it's the same code base And it's just a bag of spanners underneath.
Um, it's as if the vendors are betting that customers, um, will find it easier to deal with fewer suppliers, which until recently was true, but to your point about. Crowdstrike. I mean, um, there's going to be a lot of people saying, and Musk has already gone out and said it, I'm taking that stuff out of my tech stack.
If they take it out, if they take that out of the tech stack, we're back to, I need to get five different things to make a platform. So I just wonder if the pendulum swung all the way back. To doing proper engineering, solving proper problems and creating a category because a problem you solve is big enough by itself to have a TAM without bolting things on and making a camel out of it.
[00:15:34] Jonathan: So what are the governance responsibilities of large technology companies that profess to offer platforms to large companies and as we have seen can be found wanting?
[00:15:44] Sumir: So when we start thinking about the responsibilities of companies, right? If you produce any product today, so you produce a car or you produce any, basically any product, if it hurts people, then you have to pay them, right?
You have to pay a lot of fines, you have to pay compensation, et cetera. Why is the software industry any different? If the software people, right, if the whole industry, if they decide not to test their software, right, and actually put software out that actually damages people, how is the responsibility shifted somehow to a governmental body or someone else than the software business itself?
This doesn't happen with any other product today. And yet in the software industry, so, you know, all our sort of titans of software saying, well, this is, you know, this is what the government should do, but actually the government can't do this. Right. They are way ahead of government. They need to test software in such ways that it doesn't damage people.
They need to think about that damage, rather than throwing up the arms and saying, well, actually especially in the case of AI, for example. If an AI system makes a mistake, how is it our fault? We only build a system, but the system's doing its own thing.
[00:16:56] Paul: A lot of tech people do want to abrogate their responsibility.
The CEO of Telegram has been arrested in Paris. Again, uh, you know, there's another chap who's saying that, um, you know, don't blame the postman if I deliver a, a letter bomb. So, but at some point, you've got to stop being just the engineer software. It ain't me, Gov.
[00:17:16] Sumir: They're entirely culpable, right? I mean, if you look at, uh, A lot of the social media and the challenges that are happening with social media with young people, with politics.
I mean, of course these guys are culpable, right? And what happens is that they are all about making as much money as possible using people as a product. Well, there's a point at which using people as a product needs to stop, right? You need to sort of go, hold on, I can't abuse people like this. I can't abuse, you know, society like this.
I need to take some responsibility. And I think it's an interesting debate because most of the media is simply not willing to take on these firms because they are very, very powerful. I think many of them are much more powerful than most governments and they're also fairly consistent, right? And, and then guess what they do?
Meta goes and employs Nick Clegg, right? As their, their policy guy, right? Um, now how does that work?
[00:18:11] Paul: We've talked, we've talked about this before and the irony of all ironies is that he was the guy that said, Thou shalt not have ID cards in the UK. Uh, and yet Facebook's got everybody's ID plus a whole lot more information, and he's now their, their comms guys.
And yes, it does seem as though we're pretty aligned on this, that it's time that, uh, tech grew up and stopped just focusing on dollar signs and started, uh, thinking about its wider responsibilities.
[00:18:43] Jonathan: One thing we all want to talk about is something we've sort of noticed through our entire, uh, tech careers. It's really that tech doesn't seem to like government and it does seem to be an unspoken agreement in tech business that really government should keep its nose out of IT. The reasons you hear are often that the government is too slow moving, it doesn't understand the industry, it's driven by short term considerations and some pretty venal agendas, but Samir, I think you have a
[00:19:09] Sumir: pretty different view.
Um, as you think about technology becoming pervasive in everything, single thing we do, and if you just look at the CrowdStrike issue recently, which meant Not just 8 million computers, you know, um, had a blue screen of death, but actually that meant that people couldn't fly, right? That meant that people couldn't go on train journeys.
It meant that hospitals had to close down, operations couldn't be done. You know, that is a pretty big, major issue, right? That's a, that, that's a countrywide and a You know, global issue that governments have to get involved in. So I think governments are going to have to step into tech, into controls of tech, into fines, into, into what's right, what's wrong, and actually sort of behaviors that technology can do and not do.
Because when a system crashes, like the CrowdStrike one, and it affects, you know, eight million, uh, machines, Then, you know, our country suffered, of course, the UK government and the US government and pretty much every government in, in the Western world, you know, needs to take care of this, right? Needs to get involved.
So I think for, for the software folks or tech folks to say that government should not get involved. I think that's sort of, we are past that point now.
[00:20:36] Paul: Increasingly, these two things come together. It seems to me that we can't really stop. government and tech from becoming entwined because they just got to.
[00:20:45] Sumir: I mean, we're using tech more and more. And by the way, with the advent of AI, we're going to get 10 times the number of bits of software that we already have. We're going to be automating everything that's not been automated because it's so cheap to write software. So tech needs to be managed in exactly the same way as the government manages anything pervasive, right?
It's a effectively, it's a, there's an underlying sort of utility to tech. And that has to, whenever it is important, it has to be performant and it has to be tested and et cetera. Otherwise. Otherwise, you know, the people providing that service will need to pay fines and, and will need to sort of, you know, own up to the kinds of sort of losses that, that actually the entire country is facing, like from the CrowdStrike issue.
[00:21:33] Paul: Has things changed? Do you anticipate more government, shall we say, involvement, interaction, you know, do you think these days? The government's stepping up or is it still at arm's length?
[00:21:47] Sumir: I think there is a problem with AI right now, right? It's a hallucination. So if you can't really have an employee that gets You know, two out of five questions wrong, or even one out of five questions wrong.
And when it's a systematic error, that's just too high an error rate, right? So, yeah. So this is much worse than self driving cars. They have far more advanced right now, and we still don't trust them on our streets. I think AI has got a few years still, and right now we're in this human augmentation journey.
Not in the automation where we replace people, but there needs to be an honest debate, first of all, from the people who understand AI, but then also, I think a collaboration with the governments, because fundamentally the challenge is that the governments are going to struggle to figure out how to make sure that these systems don't damage.
Individuals or entire populations of people. And I think the only people who can really do that are the industry themselves. And like with any product industry and like with any product, normally when your product damages people, right. If it kills people, you know, whether it's a microwave or a car, yeah.
You generally tend to have huge fines. You, you know, have compensation claims, et cetera, right now in the software industry, we seem to be getting away with this, right. And that needs to change. I think the software industry needs to. Um, to basically own the testing, thorough testing of their software so that it doesn't cause damage.
And just to say, well, this thing is artificial intelligence, it's really hard to test it. Well, that doesn't really stick. I know there are a lot of people in the software industry that are really anti the EU. Let's just think about GDPR and the EU AI Act, right? GDPR is basically just saying don't misuse personal data.
Which is, feels all right to me. And the second one, the EU act saying don't build systems that misuse personal data in such a way that you manipulate people. So if you actually look at that act and look at the types of misuse, it's things like don't target groups of people and actually sort of subject them to, you know, worst treatment, right?
Et cetera. Yeah, of course you shouldn't do that. Right. So I think the, the. Where the legislation is coming from and the thoughts behind it feel right to me, um, in terms of his general sort of nature. Now whether it's actually going to be implemented properly, You know, and and what the cases will look like.
I believe it's a great start.
[00:24:16] Paul: This is not the turn I think most people would expect when we're talking about government and tech. We're actually saying this, you know, not only a role, a very valid role and, um, one that that potentially is going to grow certainly with AI for the future. So, um, Yeah, I think a lot of people would be surprised that you are advocating a greater, um, involvement from the government.
But then, as I said, uh, previously, you know, um, you seem to be aligned with Elon. So maybe that is the way forward. Well, I think we need to be
[00:24:44] Jonathan: clear here. What Samir, I think, is advocating is collaboration. Better collaboration between, between government and industry. And I think there's, um, there's too much visceral Um, loathing between the two at the moment and people need to grow up and that needs to change because the stakes aren't now so high with AI, which would you agree with that summary, sir, 100%.
[00:25:10] Sumir: I mean, when you think about. Two, 3 trillion companies, right. In terms of market capitalization, right. They have to have some sort of responsibility, right? Yeah. They can't just say our only responsibility is to our shareholders. They have to have responsibility to the entire global population. Really? They are too powerful.
To say no, that's you know, that's not our problem. So of course they're going to have to work with the government
[00:25:43] Paul: Sameer we were discussing briefly, uh a piece that appeared recently that said that um, You know, it's really hard to get into tech and uh that um, it wasn't law It wasn't accountancy and it wasn't even consulting that was the most elitist. It was tech. Um, i'm here to say it's not true I'd love to know a little bit more about your background and maybe you could start us off with like, you know, where did, where did you start your journey to being a tech entrepreneur?
[00:26:11] Sumir: Well, I certainly don't think it's true and it's certainly not my experience. I'm an immigrant. I went to actually one of the worst schools when I came to this country, when Tony Blair came into power. They actually had, had them listed as the worst schools in the country. And thankfully it was shut down a couple of years after that.
So, you know, I was very happy to leave that school and I was lucky to go to Warwick, uh, University.
[00:26:33] Paul: For American listeners, what is Warwick University? Warwick's a specific type of university. It's not your typical Red Brick. It's not Oxford. It's not Cambridge. It has, um, let's just say a technical bench, right?
[00:26:43] Sumir: Yeah. So Warwick is one of the Russell Group universities, which, which is basically, it's, it's a You know, it's sort of secondary, I suppose, right off in, in the UK. It, you know, in terms of research sort of normally in the top 10. And, um, and actually I'm, I'm delighted to say that my foundation now works with Warwick and, and Delhi university and trying to get more women into higher education.
So I've sort of kept my links with Warwick because I love my time there.
[00:27:07] Jonathan: You followed what many people would think be a normal path into IT by working in large organizations. So you could see how both IT work and how. The large organizations worked and I think the BBC was, was on that list. I think you were at Microsoft for some time.
And tell us a little bit about that part of your, your route to entrepreneurship.
[00:27:31] Sumir: Sure. So for the first few years, I worked at the BBC, Reuters, um, Lombard, which is a part of the RBC, um, Royal Bank of, sorry, RBS, Royal Bank of Scotland, and then Microsoft. The one thing I'd say is. It's incredibly valuable for young people to work at large companies.
Everything works, the organization works, the business works, of course, it's a large, you know, business, you know, people know what their roles and responsibilities are, the systems work, the tools work, they have processes. And the first few years are so formative, that once you've done that, when you start your own business, Everything's easier because you know, you know, you know what HR does, for example, you know what processes you need to do, you know what you need to do with a PO, right?
A lot of people today, you know, and, and this is a conversation I've had several times recently, actually, um, I sort of, you know, out of college, they're bright, of course, they're incredibly bright, but they want to do a startup straight away. And I'm sort of going, well, are you sure you want to do this?
Because a startup is not going to have the organizational depth. It's not going to have processes, tools, et cetera. I think for young people getting to know the basic ropes of how organizations work and how successful organizations work is actually a skill that you can't buy in an MBA. And the best thing is you get paid for it.
Right? I mean, you're, you're getting trained and you get paid for it. Take the money, get trained, and then do something else.
[00:29:01] Jonathan: It strikes me, um, in the early days that you were implicitly, if not explicitly, becoming a category creator because one of the things you were doing was reframing what customers could expect.
From a supplier, you know, not just supplying tools for Windows management tasks, but automating them and in that delivering clear benefits that couldn't be gained from existing manual category suppliers.
[00:29:31] Sumir: Yeah, absolutely. Um, I think one of the things that we found when we were at Microsoft was there was a limit to how much automation you could do because the software just didn't do it.
So actually, we had to build software that did the automation. And that's really what software is all about, right? I mean, software is about automating stuff. And so we were seeing that a long time ago. I mean, it's just a simple process of switching on computers. There was technology built into computers that you could switch them all on over the network, but no one was using it.
So we built some software that allow you to do that in a sensible kind of way. Um, but then, of course, if you've got a large building with 10, 000 computers, you can't switch them on at the same time. Guess what happens? You get a brownout. Right? They, they, a lot of the fuses go. So actually, then you have to sort of think about, okay, well now I need to phase these things in.
Um, so you have to think about automation and then degrees of automation. Uh, we've certainly, we're thinking exactly in that way. Um, and it was fun actually, it was so much fun.
[00:30:35] Jonathan: Both Paul and I always say that if you're going to, if you're going to run a business, you're going to be working very hard. So it needs to be fun because there's really no point in doing it otherwise.
We hear that even in a modest way, you might be back in the game with a small AI related startup. Um, what, if anything, can you share and what are the lessons which are now seasoned, exited and proven entrepreneur. Can bring to the table To people who might just be starting out on that journey. What does it take to run a business?
[00:31:09] Sumir: The thing is every entrepreneur is different, right? And so I know this from being a judge at the ey entrepreneur of the year for the last three years And and so i've met, you know, many many entrepreneurs um And the first thing I'd say is don't take my advice for a start, right? You know, this is your journey.
Um, but if you are going to, but if you're, exactly, if you are going to, then, um, you're clearly going to be great at some stuff, right? And actually it's going to be a bunch of stuff. You're going to be great at. And if you're not sure what you're great at, This is something you need to make sure about, right?
It's just ask people around you. And actually I find that people are incredibly honest when you just ask them that sort of open way, you know, just tell me about me. Yeah. But we, we actually often, we don't do this often enough. And this is the easy stuff for your business and your company is going to be really good at this.
Yeah. It just says, because you're running it. So the thing you've got to do is you've got to sort of make sure that you know this. We are going to be good at this and actually we are going to focus the business in such a way that we are going to make the best out of these skills. And those could be invention, which is the thing I really used to enjoy, or it could be things like relationships, right?
Which is a thing that actually I used to struggle at, right? So it depends what you're good at, right? And, and actually businesses can really blossom when you're clear about what it is that you do well. And, and then also you need to really think about. The people that you're hiring because what we tend to do is we tend to hire people that are just like us And if you looked at my exec teams Most of the time they were like me, right?
They weren't relationship people Yeah, and actually it was only when we started doing some of these personality tests that it became so obvious Right that i'm basically hiring people like me That we actually made decisions that we are now not going to do that and actually it made a big difference To the, to the, to the business, I think it made a bigger difference to our employees because suddenly they could, you know, employees have people they could relate to if they weren't like me.
Right, of course, it makes a big difference. Um, and then once you're focusing on the, the things that you're good at for the whole business, and that's basically your values, your mission and all sort of personify those, then keep at it, right? Um, there, there's going to be lots of ups and downs, right? Um, but if you think about a little growth.
Every year it compounds, right? So in the last three or four years, if you've been going for a while in the last three or four years, you've got to make the most amount of money in the last couple of years, you probably make the most amount of money. So the longer you go, this compound effect is a magical thing.
You know, I think I even Einstein's got a sort of phrase about it. And so, you know, if you just keep at it, actually the returns get bigger and bigger. I think my returns are. Fairly good. Given I started 500 pounds, I think the business is, uh, is worth over half a billion now,
[00:34:14] Paul: one of the things in conversation with you that we've talked about is, um, this, another p word purpose from our perspective is, is as, as myself and Jonathan running entrepreneurial companies, you, you know, there's a lot of scrappiness, there's a lot of doing the do, there's a lot of, as you said, ups and downs.
Purpose is a. Bigger sort of, as you mentioned the word mission, and I know that since you've exited, um, you know, you know, maybe isn't to the ground stone seven days a week as it once it was. And, you know, you're now on the other side of the journey in some respects for that first journey. Um, can you talk to us a little bit about your personal journey on purpose and where you've got to with all of that?
[00:34:53] Sumir: I always felt a purpose, a sense of purpose, you know, building tech anyway. And I think a lot of it, you know, initially it was around energy savings, even when we, you know, weren't looking at making money out of it. Um, but something we started doing very early on was we started sort of, um, supporting some kids in an orphanage in India.
And I could not believe the power of doing this across our, uh, my colleagues. Everyone signed up to this, right? Everyone wanted to support this. And it wasn't even I was, I was asking for a contribution, right? Um, and a lot of those kids, you know, we now still support. So that's a relationship that's been there for over 20 years.
Some of these kids, well, I call them kids. These are, you know, these are 20 some year old, uh, adults. Um, they just need some level of mentoring, you know, uh, they've come to an orphanage, they don't have adults to speak to, they actually don't have a peer group that is actually highly successful that can show, show them the ropes.
So we can just open those sort of doors for them. Um, it's not always been massively successful, but I, it's something that I'm. Uh, incredibly proud of actually, and, and pretty much everyone in, in money is still proud of this. That's great. And it's, it's, you know, it's, you have direct connection to about 30 to 40, uh, young people.
And then we're doing something much, much bigger because from a scale perspective, it didn't feel enough. So we started working at Warwick and Delhi universities in, in how to get more women into higher education. And when I spoke to some of the policy folks in India, they said, they just don't have fundamental research.
Now, the interesting thing is as a business person, I can take a very long term view, which most governments do. And I can sponsor research, which actually is, I mean, if you think about it, it's fairly cheap, right. In terms of, well, it's HD places and it's, it's quite a lot of money in a sense, but. If you look at the impact of policy, it's so large that actually it's incredibly cheap to actually just sponsor fundamental research.
So we've been sponsoring research for the last eight years with Warwick and Delhi universities and actually working with some of the policy folks. The biggest issue right now for women, young women in India, um, not going to college is distance. How far do they live from, from the college? And this is the first bit of our research, right?
And how do you fix that? Well, we initially thought, you know, more buses, et cetera. Well, you don't need any of that. What do you, what the research showed was the parents hadn't been to colleges and these colleges had big white walls around them. So they didn't know what happened inside. And just starting outreach programs with these colleges has been so successful that 2000 colleges in India have now signed up to our, uh, Our project, 18 universities are now participating in extending the entire project.
So, um, I'm really excited about this, right? Fundamental research to support policy seems to be our way anyway.
[00:37:51] Jonathan: This is fascinating the way that you have, you know, as an adult had, um, a series of purposes, but something tends to happen when you, you sell a business is that one moment you're living it 24 hours a day, seven days a week.
And almost the next moment, it's not there anymore. You made some, by my view, very brave decisions about what you were going to do with you to take you to the next stage. Do you want to talk a little bit about that?
[00:38:23] Sumir: Yeah, I feel incredibly lucky that, um, I had a supportive partner. I've sort of been through psychotherapy in the past, um, or a long period of time.
So I felt incredibly lucky that now I had time. Um, and actually, What no one prepares you for is how disruptive and how difficult selling your business is. So I sold a majority of my business and suddenly you're not the boss. You suddenly find that you've now recruiting another boss, right? And the actual people who are running the business and who have control on the business Uh, you know may or may not agree with you, right?
So first of all, you've lost your voice, right? You know and you have to accept this and actually they don't think you've lost your voice, right? They actually do want to listen to you, but actually they control it, right? And so but how do you cope with this? Um, so I was incredibly lucky that I had a psychotherapist Very understanding wife you should start with the wife Then psychotherapist.
And then I also got a business coach, right. And a couple of things I sort of found out of this, the first, the biggest takeaway, um, that the business coach, uh, you know, told me, and I think he's absolutely right, was just take a period of a year or two years of experimentation. Don't get involved, don't get another job for a bit, right.
If you don't have to. Um, and the reason for that is you've got to sort of, uh, he called it detoxify. I don't really think of it as detoxify because I enjoy it. a time, but you're letting go of, of what you've learned over 30 years. Um, and, and you're sort of trying to sort of find yourself. I took quite a lot of time out, right?
And I enjoyed that time. Um, but actually I realized also I, I like business, right? I like sort of being a part of, you know, building new solutions. Most of the entrepreneurs I've met are not in it to just make lots of money. That's a by product. You know, they're all trying to do something, something else.
And I found that fascinating. You know, it's a tribe that I, I like belonging to, right? This is my tribe. Yeah. Because these, these people want to change the world. Right. And something that really excited me was artificial intelligence and, and how much of an effect it can have across businesses. And so I thought about the problems that we have today with AI and basically with adoption.
And so we started a business on AI acceleration, right? How do we get around These problems that are stopping adoption of AI today. So my new firm is gate AI. So it's gate AI dot AI, and it's all about AI acceleration. Um, it's very hard to adopt AI tech right now. And I feel for the next few years, it's all going to be around human augmentation because there's one human superpower, right?
It really is a superpower, which is our ability to discern between good and bad or right or wrong. So if you combine Jenny, I with this human superpower off of being able to tell right or wrong, then very quickly you get a match made in heaven, which is huge, huge amounts of capability and good judgment.
So getting is really focused around large businesses. Um, and so pretty much any sort of business that's got hundreds or thousands of people we can help. So we'd love to talk to more businesses who are looking at AI. What sort of thinking, where do I start? Or I've tried a couple of experiments. They have not worked well.
Well, we've got a way of working, which is. Always having the human being in the loop, always making sure that we use our superpowers of judgment and actually making sure it's all auditable. You can actually see where AI is coming up with the answers. You can see the justification of those answers. So the human being can actually do what they do best.
[00:42:11] Paul: Thank you for listening. If you want to learn more, go check our blog posts on bcategorical. com.
[00:42:17] Jonathan: If you have a category issue, then we can help get in touch with us.
[00:42:20] Paul: And remember. Don't be better, be different.