Are you getting caught up in the DeepSeek chatter and missing the crucial conversations on agents?

At work, agents will be your ticket to harnessing the amazing capabilities you are hearing about. However, worth noting that there are already many 'good-enough, fast-enough and cheap-enough' model options to drive enterprise value right now!

2025: The Year of Agents

As we journey through 2025, it’s clear that agents are set to redefine the landscape of artificial intelligence in our organisations. Yet, despite their promise to boost efficiency, productivity, and automate complex workflows, many organisations hesitate to take the plunge. Concerns about potential obsolescence, uncertainties around the build-versus-buy decision, and the challenges of integrating new technology with legacy systems often hold leaders back. Still, this isn’t just another tech trend—agents are emerging as proactive, intelligent systems that can revolutionise how we work, if only we embrace the opportunity with a clear, forward-thinking strategy.

Redefining the Agent

So, what exactly is an agent? Unlike traditional assistants, which often act as simple task managers, agents are autonomous and dynamic. They don’t just respond to commands; they take initiative. They’re designed to process complex information, make decisions, and even coordinate with other agents. In essence, an agent is not just a passive tool but an active participant in driving business processes forward. If you’ve ever wished for a solution that does more than just follow instructions, then agents are exactly what you need.

Real-World Benefits Today

Imagine a world where your approval processes run seamlessly, your systems cross-check themselves for accuracy, and policies are automatically applied—without lifting a finger. That’s the promise of agents. Picture a scenario where routine tasks—like drafting reports or briefs, scrutinising documents for compliance issues, and verifying information—are automated, freeing up your team to focus on strategic initiatives. These aren’t abstract concepts; they are tangible opportunities that can transform how work gets done. By leveraging agents to perform cognitive tasks once reserved for manual processing, organisations can not only boost efficiency but also enhance accuracy and reduce operational bottlenecks.

The Build vs Buy Conundrum

I propose that one of the hottest debates we will be having in the AI space soon is the build versus buy decision for agents. On one hand, developing bespoke agents tailored to your organisation’s unique needs can offer significant competitive advantages. On the other, waiting for off-the-shelf solutions—or even renting agents from established AI providers—can sometimes yield quicker wins. Yet, there’s an inherent risk in pouring resources into a bespoke system only to see it rendered obsolete by a breakthrough from a major provider like OpenAI. Take the 'Operator' feature released just a week or so ago. While ‘wait and see’ might seem safe, it’s also a missed opportunity to create immediate value. The challenge lies in striking the right balance between making a move today and flexibility for tomorrow.

Embracing a Modular, Agile Approach

The solution may well lie in adopting a modular and agile approach. By designing agents as independent, well-defined components, organisations can swap out parts as newer, better options emerge. Think of it as building a system with interchangeable parts—if one module becomes outdated, you don’t need to rebuild the entire system; you simply replace that component. An iterative, agile methodology ensures that your implementation is continually refined, with regular review cycles that keep your technology in line with the rapid pace of innovation. This approach not only mitigates risk but also allows you to harvest immediate benefits, while staying prepared for future advancements.

Laying the Data Foundation

No matter whether you choose to build or buy, the success of any agent hinges on the quality of the data it uses. If you are responsible for any part of enterprise data, it is time to think about how to avoid being the impediment when the 'rubber starts to hit the ground'. Your unique data sources—from customer interactions to operational metrics—are the lifeblood of any agent-driven system. A robust data governance strategy is essential; it ensures that the information feeding your agents is accurate, current, and reliable. Do you feel as if we tried and failed to do this in the past and wondering what has changed now? There are multiple options to address this including using AI tools to break the back of the hard work needed to get through complexity and volume.

Looking Ahead

As we rapidly start to realising the value of agents, the message is clear: the future is dynamic and adaptive. Organisations that take the plunge now, with a focus on modular design, agile development, and solid data foundations, will be well-placed to capitalise on both immediate efficiencies and future breakthroughs. 2025 is not just the year of agents—it’s the year we reimagine the very way we work, ensuring that we stay ahead of the curve in an ever-evolving digital landscape.

Don’t let the fear of obsolescence hold you back. Embrace the change now, and lead your organisation's preparedness as the AI agents are not just a novelty, but the engine of everyday success. Don't let them remain a mystery, lift the hood and start taking a closer look. Here are 3 things you could do now:

1. Spot the Bottlenecks: Take a good look at your workflows and figure out where things are slowing down—whether it’s an approval process that’s dragging on, a document-checking routine prone to errors, or any step that’s eating up too much time or money. Zero in on that one area where an upgrade could really boost reliability, service quality, or simply cut down on risk. By pinpointing your biggest inefficiencies, you’re setting the stage for swift, impactful change.

2. Rethink Who Does What: Ask yourself who’s currently handling each step and consider how things might change if you introduced an agent into the mix. It’s not about cutting people out; it’s about letting smart software take on the repetitive, detail-oriented tasks, so your team can focus on the big picture. Think of it as redefining roles rather than replacing roles.

3. Check Your Resources: Look at what you’ll need to make this work. Do you have the quality data, the right systems, and the supporting processes and skills in place? Whether it’s ensuring your data is spot on or setting up the tech framework, making sure you have the right ingredients will help your new agent hit the ground running.

Let’s keep the conversation going. Are you leaning towards waiting to see what the market offers and what your peers are doing or start taking control today?