Well, we’re quite a niche market, in that the industry is effectively run by very big suppliers that take a very long time to change. While there are some who are sticking their heads above the parapet and trying to do things differently, there's still a certain amount of archaic processes that are just wholly embedded within the industry that are very difficult to uproot.

And I don’t want to push this too hard, but there's the feeling that everyone's kind of comfortable. Buyers need a third party, and the third party goes and secures their energy contract. Great. Why change it? There’s no benefit to the third party to spend a lot of money on trying to change and innovate, and there's no benefit to the suppliers either, so I suppose there's no real external pressure coming through.

So it’s up to the smaller companies out there that are innovating to start questioning why it is we’re still working in this way. Innovation usually comes from the smallest businesses – the start-ups. But the question is whether the big suppliers embrace working with these start-ups.

I know it does. Sadly, in terms of employment, we're a living business case of that. In the old-school format of a broker, given the amount of trade we do, at the volume we do, we should probably be a team of over 50 or 60 people, but we've only just gone from 12 to 23.

We have replaced a lot of manual work with our technology, and the idea behind it is that we deck out our business with experts who can add value on top of tech. So I think there will be some significant changes in the industry, in terms of it getting leaner. Sadly, we will certainly see job losses because technology can do the jobs of those ten, 30, 100 people.

I think the focus on new roles will very much be on value-add. So yes, I think there will be very specific new roles that will be able to add intelligence to the data where AI can’t quite cut it, or to manage all the data flows that are going through. But I believe the new roles will be very bespoke and specific. One of my background companies had over 200 people – many of whom were offshored – but they were doing very manual data tasks. Most those tasks will become obsolete.

There is a huge amount of data in our industry – dynamic data that changes every second. The industry is only really scratching the surface in terms of what you can do with that information. And you can't keep up – if you're just processing a spreadsheet, by the time you've processed it and sent it out, it's already out of date. What AI would bring is the ability to be on top of that data and offer complex analysis at any given moment to back up business cases.

The energy industry as a whole is vast and there's a huge amount of innovation going on around on-site generation batteries etc., but you need the support of machine learning or data-driven AI to really work that out and back it up. Every business is different and every business uses their energy slightly differently, so there's vast opportunity to do something very cool with the data, which is not currently happening.

I think also with AI there's much less room for human error. Our industry is packed with resources and when there's so much money at stake and you're signing multi-million-pound energy contracts, things like machine learning technology and automated data is so important to build confidence and give you as an energy buyer a lot more control.

In the broader financial sector, there is a lot of predictive modelling which we could use, as it’s just the analysis of data. I'll be honest, I can't think of any specific commodity side industry where there's been successful implementation of AI as yet that I've read about.

It’s still very, very new, there's still a lot of testing the waters. The actual independent siloed analysis of data can be done on energy data as much as it can be done on payroll data, or flight data, or any data really.

But I think there will be a short educational phase, as these things often have. The industry used to be golf days and corporate events, and some people, including buyers, loved that. They don’t want to touch something that can look to be relatively faceless, but I massively believe that it’s just an educational phase that we'll go through. Some might be afraid of the technology, but I'm very pro AI; overall I think it'll be fantastic for the industry.