Inside Valorem Reply UK: Microsoft, Agent 365, and the intelligent enterprise
Valorem Reply has recently been named by Microsoft as one of its early partners for Agent 365. Partner Shaun Laughton explains what makes the firm tick, and how it is helping clients to get the most of their agentic engagements.
Around 18 months ago, you crossed from another Reply subsidiary to Valorem Reply. What motivated your move?
It was less a move and more the natural next step in something I’d been building for a few years. I joined Reply in 2020 when they acquired Airwalk Consulting. Airwalk had always been heavily AWS-focused, and my role was to define and grow its Microsoft offering, which was fairly new ground for the business at the time. I was pretty instrumental in maturing that capability, and over the next few years, it grew to the point where, by 2024, the Microsoft side had become big enough to stand on its own. So, it was spun out into its own company and brought together with integration capabilities from another Reply business that was pure-play Microsoft. That formed what’s now Valorem Reply UK.
That’s very much the Reply model. They incubate a technology area inside an existing company and keep developing it until it’s ready to operate independently, and 2024 was our year for that. So, the decision was an easy one. It gave us the chance to grow the Microsoft business well beyond what was possible inside Airwalk, which was a really exciting opportunity.
What have been some of your highlights since joining Valorem Reply?
For me, the standout has been getting to deliver work that reaches the public and does some genuine good. Two projects come to mind.
The first is Aberdeen Protects, an app we built with Aberdeen City Council and a group of local health and recovery partners to help reduce drug-related deaths in the city, providing real-time information on where to find naloxone, guidance on responding to an overdose, and local alerts about contaminated supplies. It’s the kind of project where the technology is fairly straightforward, but the impact is anything but. The second is the Indices of Multiple Deprivation, where we architected the platform and provided the engineering to get it into production. It’s a dataset that feeds directly into national funding and policy decisions, so getting it running reliably matters well beyond the technology itself.
The thing I’m proudest of, though, is closer to home: the formation of Valorem Reply UK itself. Bringing two teams together, each with its own way of working and its own culture, is genuinely hard to do well. The last eighteen months have been a real journey, and I’ve been hugely impressed by how the team has come together into a single, cohesive unit, all pulling in the same direction. More than any individual project, that’s what I’m most proud of.
The Valorem Reply team focuses on cloud-driven data and AI transformations. What are your main areas of focus, and do you have some examples of recent work?
For me, it really comes down to enabling all the new, shiny AI technology, but doing it in a robust way. The way we approach that is by helping customers understand where they actually are on the fundamentals first: security, identity, observability, platform, integration and data. Get those right, and the AI on top can be enabled and scaled far more easily. Get them wrong, and you’re putting the business at very real, tangible risk.
A good example is the work we’re doing with a major UK regulator, which cuts right across that picture. It started with data governance and platform governance and has more recently moved into AI governance, where we’ve helped them establish an AI Centre of Excellence. That gives them a way to scale their AI use cases and drive adoption safely, rather than letting it spread uncontrollably. It’s a good illustration of the point, really. The governance groundwork is exactly what lets them move quickly and with confidence on the AI itself.
The firm’s promise is around building an ‘intelligent enterprise’. What does that mean exactly?
The simplest way I’d put it is an organisation where intelligence is built into how it runs, not bolted on after the fact. You can see it in the day-to-day. People aren’t hunting around for information or waiting on a report to make a call because the insight is already there in the tools they’re using anyway. Decisions get made faster and with more confidence, and the routine work that used to soak up everyone’s time increasingly looks after itself. The organisation gets better at learning, so something that goes well in one corner of the business spreads rather than staying a one-off.
What it really comes down to is responsiveness. An intelligent enterprise can sense what’s happening, decide and act more quickly than it used to, and keep getting better at it as technology advances. The technology is the enabler, but the point of it all is an organisation that’s genuinely more capable, not just one that owns more tools.
If you could name one of your clients that comes closest to being an ‘intelligent enterprise’, who would you single out and why?
I’d be slightly wary of anyone claiming to have fully arrived, to be honest. For a large, established organisation, it’s genuinely hard to get there quickly, because you’re working with legacy systems, embedded processes and a lot of people whose ways of working have to change. A young, digital-native business can build itself that way from the start, but then it doesn’t yet have much of an enterprise to make intelligent in the first place. I therefore think of it less as a badge a client has earned and more as a direction of travel, a target you keep moving towards. The ones I’d point to are the organisations treating it as an ongoing journey rather than a one-off project.
Two come to mind. The first is a UK regulator we work with, where we’ve built out their data and platform governance and helped them stand up an AI Centre of Excellence, so they can scale their use of AI safely rather than in an uncontrolled way. The second is a prominent UK higher education institution, where we’re working to unify their data and use AI to support student retention, improve the student experience, and keep course content relevant and applicable. Neither has "finished," but both are good examples of the mindset that gets you there: get the foundations right, then build on them deliberately.
Valorem Reply has recently been named by Microsoft as one of its early partners for Agent 365. How exciting is that for the firm and its clients?
It’s a big deal, and not just for the logo on the website. Agent 365 tackles what’s been quietly holding enterprise AI back. Once you’ve got agents running around the business, who owns them, what are they allowed to touch, and can you prove they’re doing what you think they are? Being in from day one means we’ve been building the answer to that from the start, rather than scrambling for it later. Our Enterprise Agent Accelerator leans directly on those core Agent 365 capabilities, taking a client from "we’ve got agents popping up all over the place" to something structured and auditable they can stand behind. For clients, the appeal is simple: it removes the main reason agent projects stall after the pilot, allowing them to scale properly rather than hope for the best.
Finally, you’ve been with Reply close to six years now. Your LinkedIn bio says you thrive on “continuous challenge” – how does Reply provide that environment?
Reply is a genuinely unique organisation. The technology side brings its own challenge, of course, as the rate of change keeps accelerating and what was leading-edge eighteen months ago quickly becomes the baseline. But the part I find really special is how Reply encourages innovation. Each company is small enough to innovate, test and fail fast, which is hard to do in a big organisation, yet you’re still part of a global business with more than 17,000 colleagues behind you. So you get the agility of a small firm and the reach and resources of a large one at the same time. That combination is what keeps it interesting. There’s always room to try something new, and the backing to take it somewhere if it works.
