Agentic AI - what would heading into 2025 be like without another article on Agentic AI.

We’ve been seeing Agentic AI hit the press a lot over the last while, especially as we head into 2025. I recently had the chance to build a PoC and find out what Agentic AI was all about and I think I get it now – at it’s most basic it’s:

Some sort of Agent configuration (in my case built in LangFlow).
Intelligence provided by an LLM (OpenAI GPT-4o in my case).
A bunch of tools wrapped around it (e.g. input and output tools to communicate with humans, an API tool to call an underlying piece of software and two other tools to list directories and read files).
Possibility of adding other Agents to have multi-agent collaboration.

The beauty of the solution comes as you interact with the Agent, ask it to perform tasks, which it then does by figuring out how to approach the problem, which tools to use, using them and giving you a response. What does this mean for organisations then?

What is it?

Unlike traditional AI systems that respond to prompts or execute predefined tasks, Agentic AI possesses a degree of autonomy that allows it to act independently to achieve specific goals. These AI agents can understand context, reason through tasks, and adapt their decisions based on changing conditions.

What sets Agentic AI apart is its ability to break down complex tasks into smaller, manageable steps and execute them with minimal human intervention. This capability enables these systems to tackle multi-step problems and workflows autonomously, making them more akin to human employees than mere tools.

It’s important to note that Agentic AI is a rapidly evolving field. As we speak, researchers and developers are pushing the boundaries of what these systems can do, making it an exciting area to watch in the coming years.

What does it mean from a business perspective?

Organizations that embrace this technology stand to gain significant competitive advantages. According to a recent Capgemini survey, 71% of executives believe AI agents will increase workflow automation, while 64% anticipate improvements in customer service and satisfaction.

Agentic AI has the potential to change various aspects of business operations:

Enhanced efficiency and productivity
24/7 availability for customer service and other critical functions
Scalability to handle increasing workloads without proportional increases in costs
Personalised experiences for customers and employees alike

As Agentic AI becomes more prevalent, organisations that fail to adapt risk falling behind in terms of efficiency, innovation, and customer satisfaction.

The message is clear: Agentic AI is not just a trend, but a transformative force that will, over time, redefine business operations in the coming decade. Having said all of this, the tools currently in place seem either new or seemingly early in their maturity, hinted at by versions that start 0.1 or 1.1. so tread carefully.

What do I do with it?

Here are some concrete steps to get you started:

Assess your current AI readiness: Evaluate your existing infrastructure, data availability, and team capabilities. Does your Enterprise Architecture team have Agentic AI in your road map?
Define clear objectives: Establish specific, measurable goals for what you want to achieve with Agentic AI. Define clear criteria for AI project selection with your PMO office and work with business colleagues to select pilot projects where Agentic AI can deliver quick wins.
Choose the right technology stack: Select AI platforms and tools that align with your business needs and technical requirements. Does your EAO have standards in place for AI technology stacks?
Train your workforce: Prepare your team for the AI-driven future by providing training on working alongside intelligent systems. Does your Organisational Change Management (OCM) approach help transition people to working with intelligent systems?

Start small, learn fast, and be prepared to adapt as this exciting technology continues to evolve. The future of business is intelligent, autonomous, and full of possibilities.

Agentic AI – what would heading into 2025 be like without another article on Agentic AI.

What is it?

What does it mean from a business perspective?

What do I do with it?

Further Reading