ChannelCon EMEA 2026: in gesprek met Hollie Whittles
Hollie Whittles, medeoprichter van Purple Frog Systems en Power BI Sentinel en vicevoorzitter van de GTIA-community in het VK en Ierland, begint bij AI altijd met de data: klopt die niet, dan bouw je op een kaartenhuis. Ze pleit voor een plan met steun van het bestuur, governance in kleine stappen en mensen die door AI sterker worden in plaats van vervangen. Ze vertelt over het data- en AI-handboek dat ze voor GTIA-leden mee schreef, over kosten die uit de hand lopen bij een verkeerde modelinstelling en over het risico dat AI ons sneller laat opbranden. Haar advies: investeer in je mensen, dan volgt de techniek vanzelf.
Lees het volledige transcript
Normally, don't start with introductions, but I know there's a lot that you do. So I thought it would be nice actually for our viewers if you could just start with a little introduction about everything that you do within the the AI ecosystem and landscape. Yeah. So I'm a solution provider owner. So I own Purple Frog. I'm one of the cofounders, which we set up twenty years ago now. And we do data analytics and AI solutions. I'm also the cofounder of a SaaS company called Power BI Sentinel, which does data governance.
And then I have a role with GTIA where I'm the vice chair for The UK and Ireland community. And then I help them with AI initiatives, so I spearheaded the data and AI interest group. And I also helped to create the data and AI handbook, which is now a resource in the portal. So, you know, I'm I'm all for kind of paying it forward and helping other people along the way. Yes. I can imagine because the data aspect is really important when it comes to AI.
Right? I think that's something that is often, well, overseen. That makes sense? Absolutely. It's often missed. I think in the in the mad rush to kind of adopt AI, people are forgetting to actually look at the data first. And, you know, if if your data's not right, you're building AI on top of a house of cards, and it'll all just come crashing down. Because AI will confidently tell you what it thinks is the correct answer based on the data that's got available to you. And that could be really worrying, like, if boards are making decisions on what an AI agent is telling them, but actually the data's wrong.
So before we start anything, we look at the data, the data governance, the systems, and have a proper plan and strategy before we start racing ahead and and adopting AI and building things. Yeah. So do you think that's also the pitfall that a lot of companies make or a lot of companies in the MSP ecosystem make that we just got started without actually having a plan or thinking what we're gonna do and how that's gonna look like and what is needed in order to... Yeah. I mean, the plan needs to cover many different elements.
It needs that board level buy in because, you know, what are we doing? How do you know what your employee is signing up and creating? You know, what data are they accessing? Sensitive data, could be building AI agents and models and exposing data. It's also kind of the human element to this. Like, your staff might be fearful of AI. Like, if you're an MSP and you start automating things, you know, are they gonna be losing their jobs if you start bringing in AI into call centers and ticket handling and, you know, what's the impact on the people?
So the people have to be part of the journey. They have to be part of the strategy. And for us, we very much recommend investing in the people and the skills that they need. AI should not be replacing them. It should be enhancing them. I think so. One one interesting thing, was at a conference a couple of weeks ago and somebody said it's about the skill set, and then the tool set. Oh. Right? So actually starting with the mindset for change, but then also the people.
Do you think that... Also, the people are often underestimated in this whole process of AI and adoption. Yeah. Absolutely. They are definitely and they're not... Their their kind of thoughts and creativity are not taken into account, and sort of solutions are signed up for and bought before you've even spoken to the to the people. So we were bought into... It's a big manufacturing client, actually, because because we can work across any vertical. But anyway, they they had us into their manufacturing plants and and they were talking to us about, yeah, we want to bring in AI to get rid of all of these people.
And then they have these people talking to us about their jobs and how AI could work for us. We were, like, really, like, oh my god. Are all these people gonna go because of this? And you could see the fear in those people's eyes that, well, if I tell them how this process works, am I gonna lose my job? And and I think that can... That permeates across to every kind of sector, but especially in, like, MSPs in the digital space. We we need to have that strategy before we start racing ahead.
So how do we enable people actually to drive AI adoption further in the industry? I think it's about that knowledge sharing and attending events like the, you know, channel economy. There's been so much talk about AI, but there's been lots of useful information about it. So I've just been at a session run by Elka Gerhardt. So she was talking about the neuroplasticity and the and the neuroscience behind AI. And actually, you still need that human oversight. You can't just race ahead and believe whatever the AI is telling you.
And and the human creativity still needs to be very much part of that equation. Yeah. But something that she touched on it on it was that with AI making us faster and faster and faster, does that mean that we're gonna burn out quicker because, you know, we're doing so much quicker? And I thought, I hadn't actually thought about that. You know, we're building all these, like, a... AI agents' assistants and we're getting faster and faster and faster. Does that mean they're actually burning out quicker? So there has to be a plan for all of this.
And I keep coming back to the strategy and and the people very much should be at the heart of that. I think it's interesting because I think we all we all keep talking about how AI is gonna make us more efficient and more productive. But the question is, what do you actually do with that extra time? Does that mean you're gonna do more work? Or are you actually gonna use the time more creative or more? Yeah. Exactly. So I've signed up to something in The UK that we have called Tech Time Out.
And every day, once a year, it's called Tech Time Out Day. And and for all of us running digital MSPs and vendors and distas, whatever you are, you take an hour out and you all sort of sign up to this digital mental health and well-being. And I think we need to see more of that because AI is gonna make us burn out faster unless we start looking after our own mental health. And we need our brains to be creative because the AIs can't think in that way.
It needs that human oversight and guidance because it will narrowly go off and do whatever it thinks is right. But that doesn't mean it's got the moral compass. It hasn't got our years and years and years of experience and and that ability to look at something and go, actually, that doesn't feel right, you know, it should be that. And an AI doesn't know that. And there was a great story about an AI that had been bought in. It was an agent and it was looking at how it could make pizzas better.
And it found this solution to make the cheese stick on a pizza with super glue because it had found an article online in Reddit that someone had written as a joke years and years and years ago, but it was confidently saying this is how you can make pizzas better because it had found this article and it deemed that as a as a, you know, a source. Yeah. I think that's very... You always need the human in the loop. Right? Need the human in the loop. Absolutely.
So you mentioned the AI handbook. Can you actually explain a little bit more about that as well? Yes. So that was a really interesting journey that we went on because we were worried about people signing up for like AI readiness tools and all of a sudden they've got Copilot and then they've bought this tool off the shelf and it's replacing Zendesk or whatever. Yeah. The tool is actually starting at the beginning. Like, how do you come up with a strategy? How do you look at your data?
How do you look at the governance of the data? How do you analyze all the different systems and the sources? What reporting are you trying to do? What decisions are you trying to make? What... What's the pain point? What's the reason for all of this? Yes. And then how do you actually build a solution and the strategy around that? So that basically the guidebook takes you through all of that different stuff and then it ends up with AI at the end. Because you can't do the AI part till you've done all the other steps.
But it is what we do. Right? Yeah. I mean, we just get started with co pilot. We just we just have some licenses or we have some or whatever. Yeah. And we just get going without actually thinking about what it is that we actually wanna achieve. Yeah. Exactly. So that resource is now sitting in the GTIA member portal. And actually, you know, it was a UK and Ireland thing that we wrote, but actually, it's global. It doesn't matter what country you're in. You can pick that up and use it.
And it's got loads of different references in there and sources, and it's vendor neutral so we're not, like, kind of pointing anyone on one particular technology. So, yeah, it's a... It's... It was a really empowering thing to be part of. I'm really proud of it. So how do you... Where are we from a from a governance and... Where are we from a governance perspective with AI when we look at the market? I think governments are now to take note because I feel like AI has certainly raced ahead in the last year.
You know, it's going faster and faster and faster, and we haven't actually got the governance behind it. And certainly, some of the, like, the reporting platforms, like, don't... You know, when they rolled out stuff like Power BI and Microsoft to try to democratize users of Power BI to create your own visual reports. But how do you know, as the owner of a company, what your users are creating in those reports? And there's a great example of a client that we worked with, and they just wanted to create an HR report to look at the hours that people are spending on things.
But the problem was in the data model behind the report, they pulled in the whole HR database. So actually, you could have filtered in Power BI and added in salary, disciplinary, and all of a sudden you've got all this stuff that you didn't expect. All information. Yeah. Yeah. So the governance part of this is absolutely key, and that's just a reporting example. It's the same thing with your AI. That's why you need the guardrails around it. Yeah. And people are using AIs to create AIs, which is fine, but then, you know, what are the guardrails around that?
What are the ethics and the morale morality around that? So I always would say you need to use an AI expert. You can't just go have your staff going off building whatever they like. But how do we how do we keep that practical? How do we make sure that that's like the plan side of things and the governance side of things that that that doesn't really become that big paper tiger. Right? So how do we how do you keep it practical? I think it's small bite sized chunks and having it accountable through board level meetings, having it part of meeting minutes and strategy meetings and rather than just allowing the users to go off and raise their hand doing stuff, they have to be held accountable and you have to embed that cultural change and that process change to be able to track it.
I think that's interesting. Cultural and process change, you think that's very complicated for for a lot of companies. Yes. 100%. And also, there's there's the cost element to it. Yeah. So, you know, if you signed up to, like, Claude, for example, you know, if you've if you've got it on the wrong model setting, all of a sudden, you could be blowing through all your credits and spending thousands and thousands of pounds. All of a sudden, like, your company credit card's maxed out and, you know, there needs to be that oversight and that governance and the control of the AI within the organization.
I do think that there's also a risk there. Right? Because, I mean, we're pushing and we're moving towards AI and I mean, it's it's here to stay. Right? So I definitely feel it's not a hype, but it's here to stay. But we're pushing that way. But I think from a financial perspective, it's it's going to have an impact as well because we're going be heavily depending on AI. Yeah. 100%. Do we balance that? That's a really good point. And, you know, we do... Our government pushes out surveys to us and and it asks us some feedback things.
But one of the questions it never asks is about the cost of cloud, the cost of AI, and, you know, it's all about environmental costs. Yeah. But actually, I think this is where we need to look at. Well, the big companies, they're just pushing their fees up and up and up. And, you know, once you're sucked into something and you're signed in, that's it. Your cost... You'll you'll be so far down the rabbit hole that it's gonna be very difficult, very challenging to actually get out of it.
Exactly. So, you know, you have to have that board level governance, that strategy. I... You know, it's as simple as that. And I think if if people haven't got that, they need to just say stop. We need to document this. And I know that sounds old school, but literally, there needs to be a plan. Yeah. So how important is community in this whole AI era? I think community is absolutely paramount because you need to learn from your peers, you know, what are they innovating and what are they doing in their businesses, what what mistakes have they made, you know, have they adopted things where it's actually opened up a cyber security risk, you know, what what are people experiencing in the community and the channel.
And that's why events like this are great because you've got all those people, all those brains in one room together. And we're inspiring each other, but we're also sharing that knowledge and that best practice. So if you were able to give one key takeaway to the people watching us, what what would that be? One piece of advice. I think for me, it's invest in your people because then the technology will just follow. So for me, people at the heart of everything. Thank you. You're welcome.
