In 2024, AI has led the way as a popular business investment for so many, from SMEs to multi-national corporations. In Endava’s Emerging Tech Unpacked Report 2024, AI and generative AI were highlighted as the top priorities for those surveyed. Yet, in an era defined by constant invention, AI’s potential continues to evolve.

In the past, AI has relied on constant human input to ensure expected outcomes and decrease the possibility of mistakes, from misinformation to hallucinations. In doing so, its application has been constricted to short-term tasks, instead of scaling its automation for the long-term. But with the advent of agentic AI, also known as multi-agent AI systems, this narrative is changing. This technology doesn’t need the level of human intervention that traditional AI does, making it easier to scale in a way that suits business needs.

But how can businesses ensure that investment in agentic AI suits their long-term goals? What elements are necessary for a successful AI transformation?

Steps to Success

No matter what kind of AI corporations are looking to implement, they need to ensure that they have a solid data-based foundation to support the first phase of adoption. Worryingly, many UK companies are unprepared; the National Bureau of Economic Research revealed that 57% of companies are using software with severe security vulnerabilities. Organisations relying on outdated systems, and reluctant to modernise, will struggle with AI integration. To truly innovate, they must first build a secure infrastructure capable of feeding data into scalable, automated centrally located solutions.

At a basic level, businesses can integrate AI capabilities with day-to-day tasks, automating administrative duties like generating reports or summarising text. While this undoubtedly offers speed and cost efficiencies, the real value of AI emerges when it takes on more responsibility.

For those wary of this shift, it’s important to remember that the highest value AI delivers stems from its ability to act autonomously. As well as summarising information, as it had done before, AI is able to advise on next steps, work without direct human intervention, and facilitate sufficient data sharing with large language models (LLMs) and cloud-based platforms. As the need for people to set the solution up for singular tasks falls away, AI, and more specifically agentic AI, can take on tasks that were previously only undertaken by humans and scale faster than has been done before.

Consider the example of a bank account application process, which typically involves multiple document submissions and checks. Agentic AI can streamline this process, supporting users with minimal human involvement, significantly reducing delays and increasing customer sign-ups. While it is still important to have a human in the loop for quality checks, each step of the process no longer needs to be performed by members of your staff.

But how exactly does it work?

In a multi-agent system, each agent is assigned a role and the data they need to complete their tasks. When described, it sounds very similar to a human team, yet this is all done with minimal human oversight, enabling workers to add higher level value elsewhere in the company. These ‘agents’ have the capacity to communicate between themselves, react to unexpected change in the data or processes, and understand the context of their tasks to ensure any decisions made lead to the optimum outcome. They can also “raise their hand” when an agent hits something in a process that they are unsure of and can seek guidance from an actual member of staff.

While this technology is not as prominent in the business world as a range of more short-term AI solutions are, business executives should have faith that agentic AI can automate workflows without concern. Through their ability to perform repetitive tasks in a shorter time than their human counterparts, they can ensure quick completion, driving substantial scale efficiencies in the long-term.

In turn, those employees who would otherwise spend most of their working hours completing mundane, administrative tasks are able to take on challenges that require creativity and analysis. When able to see their growing potential in the business, and feel that their contribution is being recognised, this will should lead to a growth in worker retention and happiness too.

Adoption by Highly Regulated Industries

While the excitement surrounding AI is understandable, businesses, especially in highly regulated sectors such as finance or health, must remain cautious about how they manage customer data. Strict legislation governing these industries has traditionally made them wary of embracing innovation and adopting AI tools for automation. This is where agentic AI offers a district advantage.

Those working in highly regulated fields must be aware of which data they’re using and how to meet compliance regulations. Although agentic AI reduces the need for human oversight, this doesn’t mean it compromises compliance. AI systems usually function as ‘black boxes’, meaning that their processes are hidden from the user, who only sees the beginning and end output. However, with agentic AI it is possible to build a multi agent system where all the data being used is captured, logged and displayed in a user-friendly manner.

Essentially, through a data-first approach, users should be able to view the information that has been captured by AI to understand how decisions have been made. When they can see all the data and various processes used, it becomes easier to both avoid AI related pitfalls, for example hallucinations, and to prove regulatory compliance.

Looking to the Future

Agentic AI promises to help corporations automate processes, saving time and money, and scaling at speed. There will be people-centric benefits too, with employees able to spend less time on menial tasks and more time bringing value and creativity to their place of work. This is fundamental to having a competitive advantage, in the realms of revenue growth, innovation, and talent.

By following a data first approach, businesses can ensure that even with reduced human instruction, their processes remain smooth, secure, and compliant with local legislation. This approach sets the stage for a future where AI not only transforms business operations but also empowers people to reach their full potential.

Joe Dunleavy, Global SVP, Head of AI Pod at Endava.
David Howell

Dave Howell is a freelance journalist and writer. His work has appeared across the national press and in industry-leading magazines and websites. He specialises in technology and business. Read more about Dave on his website: Nexus Publishing. https://www.nexuspublishing.co.uk.

Recent Posts

New EU Law Expands Digital Resilience to Third-Party Dependencies: What is the Impact on Businesses

The EU’s Digital Operational Resilience Act (DORA) sets new standards for financial services, emphasizing digital…

1 month ago

So long, SaaS: Klarna is right, DIY is the Future for AI-Enabled Businesses

Klarna’s bold decision to abandon SaaS giants like Salesforce and Workday signals a major shift…

1 month ago

Demystifying AI Models: How to Choose the Right Ones

Large Language Models (LLMs) have revolutionized artificial intelligence, transforming how businesses interact with and generate…

3 months ago

Beyond CISO Scapegoating: Cultivating Company-Wide Security Mindsets

In the evolving cybersecurity landscape, the role of the Chief Information Security Officer (CISO) has…

3 months ago

Three Key Considerations for Companies Implementing Ethical AI

Artificial Intelligence (AI) has grown exponentially, transforming industries worldwide. As its use cases expand, concerns…

3 months ago

The secrets to Developing a High-Performing Data Team

Building a high-performing data team is key to leveraging data for better decision-making. By balancing…

3 months ago