The Advent of Agentic AI
It all started on Nov 30, 2022, when Open AI launched ChatGPT, which became viral in just a few hours. As the world was getting their head around this new term, enter Generative AI. And then followed Agentic AI. This left many wondering what the difference between Gen AI and Agentic AI is, since both had ‘AI’ in common.
Gen AI and Agentic AI – Same Roots Different Leaves
In the recent years, Gen AI has caught everyone’s attention because of its ability to create text, images and code using natural language and simple prompts. These are more human-like and the true potential across multiple industries is being recognized rapidly. Here is the interesting bit – while Generative AI or Gen AI can be seen as the system that creates the outputs, Agentic AI is what delivers the outcomes. Also, while GenAI excels at generating text, images, code – Agentic AI goes much further by adding reasoning, making decisions or executing the tasks.
The Real Business Value of the Two AIs
Generate AI can help in creativity, improve efficiency of business processes across various business functions like marketing, services and product development.
At the same time, Agentic AI is emerging as a game changer for automation, customer support and operations.
Where can Agentic AI be Applied?
Agentic AI can be used in areas which require a certain amount of reasoning to make decisions for specific goals. Slowly shifting from the “conversation” mode to a “creation” mode and finally to “what can be achieved” mode.
Let us take an example of large deals which have a lot of run and change operations in the scope of work. In these scenarios, an Agentic AI solution can be leveraged in creating the run and change agents. All of these can be specialised in the way of operations and can achieve the so-called automation, reasoning, decision making and execution at the end.
The Real Outcomes from Agentic AI
There is enough interest generated in Agentic AI, to make enterprises set aside budgets for it. The organizations have experienced productivity improvement as a primary benefit of using Gen AI. Now, however, it is Agentic AI, the next blue-eyed boy(/girl) who has been passed on the baton, to help create meaningful outcomes, reduce costs and deliver benefits at large. With Agentic AI, businesses have been able to visualise how they can do achieve at enterprise scale.
However, Nothing Comes Easy – Cyber Security Concerns
In discussions with multiple clients, we see that they are slowing down the decision process as there is a genuine concern on Gen AI usage and its obvious problems of hallucinations. This can be overcome by use of Agentic AI which also brings human in the loop at the right moment.
Cost of Implementation
The cost of implementation is also something that worries most organizations. They have started implementing guard rails and usage limits now, started usage of copilots to limited groups on experimental basis. The reason for lesser adoptions is because of companies themselves spending very cautiously on licenses or tokens.
Agentic AI to the Rescue
Overall, the shift to Agentic AI is happening at a rapid pace. Companies should start asking employees to start experimenting with simple agents and building end to end automation workflows. There by start the revolution internally first by greater adoption for routine tasks and then they gain confidence and explore their client/customer facing implementations.
What’s Infosys Doing?
As part of Infosys Topaz, the company has developed Topaz Fabric to define and lead the Agentic AI space and support the creation of operational agents. It is a composable enterprise AI platform that layers data, models, and AI agents over existing systems to automate IT operations, accelerate modernization, enhance quality engineering, and strengthen cybersecurity—bringing together digital, AI, and human workers to deliver faster, more cost‑efficient, and resilient services.
