The Age of Agentic AI: Shaping the Future of Business Operations

Recently, Mark Purdy for the Harvard Business Review has firmly stated that we are entering the “era of agentic AI”. Unlike past waves of automation, this one doesn’t just execute - it decides. A glance at LangChain’s State of AI Agents report (2024) makes that clear. 

LangChain, a leading company in LLM-based applications, surveyed 1,300 entrepreneurs to understand the state of agentic AI for businesses today, and where it's heading. It emerged that

• More than half of the respondents were already using agents in their daily tasks;
• 78% of those who weren’t, planned to integrate them soon into their company resources.

This made me wonder: have we missed something, and is the era of agentic AI already here?

DataHub.io is here to provide a clearer picture. We conducted a cross-sectional research of how companies implement agentic AI to maximize results, today. We will go beyond a simple introduction to this new frontier: we will look at four concrete examples that could influence your view of technological integration in business.

What is Agentic AI?

Let’s face it: we’ve just gotten used to the innovations brought by the expansion of AI, fascinated by the power of algorithms and their ability to process huge amounts of data in tiny amounts of time. Yet, the exponential growth of this technology has only just begun to amaze us.

Agentic AI, the new frontier of IT, could once again change the world we know – from healthcare assistance to care jobs in health services. How? 

Unlike traditional AI tools that respond to prompts, agentic AI is built to act. These systems operate autonomously, make decisions, and pursue goals with minimal human input. In essence, they’re not assistants — they’re agents.

Generative AI informs. Agentic AI performs.

AI agents are digital workers capable of solving problems autonomously, pursuing pre-set goals rather than responding to prompts. 

Unlike generative tools like ChatGPT, agentic systems:

• Gather data to understand real-world context;
• Make decisions based on objectives;
• Adapt autonomously, minimizing human oversight.

The technology on which AI agents are based includes large language models (LLM) to process information and reach the given objective. However, they are tailored for tasks far beyond text and can interact in a human-understandable language. Think for example of   Moveworks’  Power Design’s copilots: they use conversational data to interact with employees, proactively managing their IT resources and workflows.

Now widely regarded as one of the greatest IT innovations, agentic AI has roots that stretch further back than one might expect. The year is 1997, and a historic match is about to take place: legendary chess champion Garry Kasparov faces Deep Blue, IBM's groundbreaking software, and suffers an unexpected defeat against a machine. From that moment on, robotics and automation in technology would evolve at an unprecedented pace.

Some even trace the true roots of agentic AI all the way back to the brilliant Alan Turing, a scientist who, in the very challenging times of World War II, revolutionized technology and computer science forever.

So why are we only talking about agentic AI now?

The golden age of agentic AI

While the tech might feel new, the ambition behind agentic AI goes back decades: to Alan Turing, to Kasparov’s defeat by Deep Blue, to every moment humans dreamed of offloading cognitive work to machines.

What has changed then? Scale. Access. Usefulness. 

The technology is no longer theoretical. It’s being embedded into core business processes, from email triage to compliance workflows.

We are talking about the closest thing to autonomous robots among all the technologies that have come so far. If you think this is science fiction or technology for the few, however, I’ll stop you right there.

Agentic AI applications are expected to significantly impact the business world: Gartner predicts that, by 2028, one out of three enterprises will include agentic AI in their software applications, with at least 15% of routine decisions being made by agents.

This isn’t just a forecast. Although few know it, it’s already happening, as seen in LangChain’s 2024 survey, where over half of respondents reported using AI agents in their day-to-day work.

Thanks to their ability to understand context, act independently, and learn over time, AI agents are especially effective at:

• Manage routine tasks much more easily and in a less costly manner;
• Automate frequent tasks such as provisioning and troubleshooting technical problems with company software.

To top it off, one possibility that’s becoming increasingly popular is combining different AI agents, creating so-called frameworks to perform very complex tasks. In 2024, Arize AI co-founder and PhD candidate Aparna Dhinakaran suggested that many new frameworks would soon dominate, becoming capable of writing our emails and booking our flights, among other things.

The frameworks often work hierarchically, with one being trained to understand the goal to be achieved, the context, and the best AI agent among others to carry out the task. This type of technology is so promising that well-established organizations have already dived into it, including. 

Microsoft is one great example of this new trend. The company started to collaborate with OpenAI and invest billions in advanced AI models in 2019, coming in 2025 to integrate it across its platforms through CoreAI. It has also recently announced the integration of eleven new AI agents into its security tools. Amongst the tools implemented by Microsoft, we find: 

• AutoGen: a classic system for orchestrating multiple AI agents; 
• Semantic Kernel: focusing on understanding context to develop useful real-world software;
• Alert Triage: to assist experts in quickly understanding the context and best strategies to address each security alert;
• Phishing Triage: meant to distinguish between real and false threats. 

Practical real-life business applications

Let’s delve deeper into the real-world applications of AI agents and why CEOs are opting for these technologies, following a strictly data-driven decision-making process. 

Human Resources management

Data and communication management within companies has become a real “tax on integration”, said Fabio Pascali, Regional Vice President of Italy, Greece, and Cyprus at Cloudera. As reported by Personio, a study by Kienbaum unveiled that “HR staff spend 39% of their time on administrative tasks”, reaching a total annual salary expense of almost €20,000. 

Power Design’s HelpBot, which received the Service Desk Institute certification, solves the problem by focusing entirely on proactivity. The agent interacts spontaneously with employees to optimize their work by pointing out critical issues related to the management of technological resources and offering immediate solutions.

Data management

We can never stop underlining how much of a hurdle data management can be for businesses. But repetita iuvant: while 9/10 companies were investing in big data and AI already back in 2021 (source: HBR), employees still spend an average of 1.8 hours per day searching for information (source: McKinsey Report). 

The Business Insider revealed that the Big Four (Deloitte, EY, PwC, and KPMG) have all invested in agentic AI to automate data management, including financial aspects such as tax compliance, to reduce operational costs and free up human resources from tedious tasks. Deloitte, for instance, is implementing Zora AI, built with NVIDIA, to: 

• Implement the data analysis cycle from both structured and unstructured sources; 
• Apply the models to draw meaningful insights and trends, also translating them into human-accessible formats;
• Cooperate with other agents to follow more “complex and nuanced workflows”.

Care and health

Having compiled an extensive pharmaceutical data collection, DataHub.io understands the critical role figures play in healthcare. This is especially true in countries with expansive welfare systems, where hospitals face ongoing challenges in accommodating and treating all patients, a reality highlighted in the WHO's 2024 report on the declining quality of care in the European Region. 

In this context, as Brian Gormley reported for The Wall Street Journal, private health companies are already leveraging agentic AI to streamline the enrollment process for clinical trials, automating for instance the identification of potential candidates. Grove Trials is implementing an agent to manage also communication with participants and to reduce the costs of operations. 

Cybersecurity

According to Check Point research, global cyberattacks increased by 20% in the second quarter of 2024. AI is making them increasingly powerful, allowing them to quickly learn from failures and come back stronger than before.

Linking back to the insights of a group of UK people in 2013, AI is now also implemented by Dark Trace to recognize any abnormal behavior. It specializes in protecting enterprises and industrial immune systems, detecting and solving minor threats while automating defense against previously encountered and solved attacks.

The future of agentic AI and businesses

The rapid expansion of agentic AI is reshaping industries, offering unprecedented efficiency and automation. However, as businesses increasingly integrate these technologies, critical questions emerge: how do we regulate and govern AI agents to ensure ethical deployment? How can companies strike a balance between automation and the human workforce?

While concerns about job displacement echo fears from past industrial revolutions, the reality may be different. Rather than replacing employees, agentic AI is proving to be a powerful tool for augmenting human capabilities freeing professionals from routine, repetitive tasks and allowing them to focus on higher-value work. As Deloitte highlights, this shift is not about replacement but about optimization, enabling businesses to enhance productivity while maintaining a human-centric approach.

The coming years will determine how well businesses, policymakers, and society at large can harness the full potential of agentic AI while addressing its risks. 

One thing is certain. The era of agentic AI is not just on the horizon: it is already here.

The question is no longer if you'll adopt it.

It's how fast you'll adapt to a workforce where valuable robotic contributors uplift everyday human efforts.