Organisations' demands of their IT infrastructure continue to evolve at an unprecedented rate, with the drive for growth and innovation needing to be balanced against the need to maintain cost control, visibility, and - crucially, cyber security. Numerous solutions have emerged in response to these challenges, but one of the most pivotal questions organisations must answer is whether to host their critical data and applications on-site, or in the Cloud.
Let's explore the respective advantages and disadvantages of both approaches, and then consider whether the increasing sophistication of modern workloads demand a new approach...
The retail playbook has been fundamentally rewritten. Customer journeys are omnichannel by default, IoT sensors are now omnipresent in both warehouses and shop floors, and AI is moving from pilot to P&L at an unprecedented pace. And the results are already proving transformative:
But in the race to access all these potential benefits, the winners aren't the ones with the flashiest demos – they're the ones with a rock-solid digital foundation that lets AI and IoT platforms scale safely, securely, and intelligently, store by store.
So, from Exponential-e's vantage point across cloud, connectivity, cyber, and communications, and our ongoing conversations with top retailers across the UK, here's what "AI/IoT-ready" actually means for the sector, and how we can begin laying those all-important foundations…
IoT and AI thrive on low latency and high availability, particularly when Point of Sale (PoS), inventory, and computer-vision workloads are increasingly interconnected. That means the underling WAN stops being a cost line and becomes a growth platform. Frictionless shopping experiences, incorporating queue-free checkout, real-time offers, and dynamic pricing, depend on fast, reliable data flows at the edge.
Software-defined networking, built on a private VPLS core, makes this practical at scale, offering centralised control, application-aware routing, seamless use of diverse access (i.e. ethernet, 4G/5G), and integrated security. Beyond the immediate operational advantages of avoiding hairpinning over the public internet and low, predictable latency, such networks offer the scalability and agility needed for pop-ups, seasonal peaks, and new store openings, where day-one uptime and policy consistency are required.
This should be complemented with enterprise IoT/M2M SIMs that deliver multi-carrier access and centralised control for store sensors, handhelds, lockers, smart signage, and similar devices.
AI-assisted retail is a hybrid sport: heavy training and data engineering in the Cloud, instant inference and control at the edge. To this end, retailers pursuing "always-on", augmented stores are converging 5G, IoT, and AI with edge compute to deliver truly personalised experiences in the moment, not hours later. This next-gen local processing, with edge computing implemented in every store, delivers a seamless PoS for customers, while simultaneously optimising staff's efficiency and reducing backhaul costs.
In the longer term, centralised data platforms and AI services can crunch multi-store telemetry for demand forecasting, replenishment, and customer analytics, offering a rich stream of actionable insights that enable reduced energy usage, automated restocking tasks, and smoother labour scheduling - immediate, powerful operational wins.
These capabilities can be developed into a standardised model and then be deployed, managed, and scaled consistently across new sites as retailers expand their operations. It's no surprise that multiple European retailers are already doing exactly this to not only protect their immediate margin and availability, but also accelerate their future growth plans.
Retail IT estates increasingly span POS, e-commerce, click-and-collect, and IoT devices. However, more devices and more data mean an increased attack surface, particularly when it comes to customers' payment data. As a result, robust security must be embedded in the design of all systems, platform, and processes, not bolted on later. Forward-thinking retailers are already rolling out this 'secure by design' approach, building customer trust through multi-layered, PCI-DSS-ready security ecosystems that allow for continuous monitoring and intelligently automated policy enforcement.
Secure Access Service Edge (SASE) has a key role to play here, converging network and security in the Cloud and offering numerous pathways to establishing identity-centric access, micro-segmentation of IoT devices, and uniform policies across stores and partners. Even with thousands of distributed end points, all this can be accessed through a single pane of glass - a "single source of truth" for all networks, devices, and workloads.
AI and IoT in retail aren't separate projects; they must be treated as fundamental parts of a single, software-defined platform that reaches every shelf, sensor, and checkout. Build the network and edge right, wrap it with zero-trust security, and connect it to a governed data and AI backbone, then scale and optimise what works.
If you'd like this distilled into a tailored blueprint for your own estate (i.e. current stores, formats, and use-case priorities), we can map the stack, identify quick wins, and sequence the roadmap to outcomes, with everything overlaid by a single SLA, as a fully integrated service. Get in touch to discuss your own AI and IoT goals and let's make sure you're building on the right digital foundation!
Software Defined WAN, or SD-WAN for short, is the new big thing in business networking. Everybody's talking about SD-WAN, and about what it can do for businesses. Well, there's no doubt that SD-WAN can do a lot for your business; in fact, we'll be talking about exactly that in part 2 of this 3 part blog series. But before we do that, we need to talk about what SD-WAN can't do. Right now there's a popular misconception among businesses concerning SD-WAN - a misconception fed and sustained by headlines and marketing hype - that could lead them to take damaging shortcuts in incorporating SD-WAN into their business.
Having returned from the BSA Conference 2025, it was clear that building societies are navigating a critical juncture - balancing their relationship-led service models while addressing the urgent need to modernise outdated systems and improve operational efficiency. Given the current geopolitical climate, and the increasing pace of innovation, developing a strategy that not only solves immediate challenges, but offers sufficient scope to tackle future issues, can often seem like a moving target.
Secure, scalable IT infrastructure for one of the UK's leading IP attorneys
World-class IT that supports the seamless delivery of legal services depends on infrastructure that is both highly available and highly secure, designed with the sector's rigorous security and compliance requirements in mind. To ensure their infrastructure provides the very highest standards of security, performance, and availability, with a consistent experience for staff across all their branches, J A Kemp selected Exponential-e as their trusted technology partner.
Exponential-e provides the underlying connectivity between J A Kemp's three offices across the UK, supporting seamless communication and collaboration between more than two-hundred end users, secured by a centralised firewall. To further optimise the resilience of their IT infrastructure, their primary on-premises servers are complemented by dedicated backups, stored in a dedicated private Cloud in a colocation environment, with Exponential-e's data centre teams managing the hardware and firmware layers, including ongoing testing, while J A Kemp's own IT team retain full control and visibility of their VMs.
With this as the foundation, a dedicated account team work closely with the internal IT team, providing proactive monitoring and alerting to ensure any potential disruptions can be identified and resolved, with zero effect on day-to-day operations.
At the same time, J A Kemp's account team work closely with them and multiple specialists from across Exponential-e to ensure new opportunities for enhancement and optimisation can be identified and acted up as soon as they arise.
Our account team haven't just been proactive in terms of renewing what we've got. They've also helped us validate what we need and ensure our infrastructure continues to evolve as the business does.
Hemel Vaja
Head of Technology,
J A Kemp
From Ambition to Enterprise Execution
Building the Foundation for Scalable AI
Turning AI into Real Operational Impact
Scaling AI with Confidence and Control