What is agentic transformation and why it matters

The term agentic transformation describes a fundamental shift in the deployment of artificial intelligence within companies. It's no longer just about chatbots that answer predefined questions, nor about simple automation of repetitive processes. Agentic AI means autonomous software agents that independently reason, make decisions in ambiguous situations, and execute complex workflows with minimal human oversight.

The concept gained attention thanks to a recent column in Daily FT, where Salesforce's regional vice president for South Asia outlined a concrete vision for Sri Lanka. His argumentation, however, has universal validity — from Colombo to Prague.

According to global research by McKinsey, only 1% of companies are fully mature in AI adoption. Salesforce also found that 93% of workers worldwide do not yet fully trust the outputs of artificial intelligence for work purposes. This "trust gap" is, according to the author, the main barrier that companies must overcome if they want to truly benefit from agentic AI.

How agentic AI is transforming key sectors

The Daily FT column identifies three areas where autonomous agents will bring the greatest impact. These are sectors that are also relevant to the Czech economy.

Financial services: From "if-then" to autonomous trust

The Czech banking sector has already undergone a digital transformation — mobile banking is the standard and open banking (PSD2) has opened the door to fintechs. The next step is autonomous trust. Agentic AI can analyze a customer's entire financial life — from investment history to spending patterns — and offer personalized financial advice in real time. This is not just about private banking for the wealthiest clients, but about an accessible service for the broader population. At the same time, agentic systems can handle KYC processes, loan approvals, or international trade finance with round-the-clock efficiency.

Healthcare: More time for patients, less paperwork

Czech healthcare struggles with chronic staff shortages and administrative burden. AI agents can take over the "detective work" of diagnostic triage — going through medical histories, lab results, and imaging examinations and preparing materials for a doctor's decision. At the same time, they can coordinate follow-up care, monitor adherence to treatment protocols, and ensure no patient falls through the cracks. The human remains in the decision loop — the final verdict always rests with the doctor — but the agent frees up their hands for what they do best: personal contact with the patient.

Education: A personal tutor for every student

Czech education faces the challenge of individualizing instruction. Classrooms are diverse, every student learns at a different pace. Agentic AI can function as an always-available study companion that adapts in real time to the pace, language, and context of each student. At the same time, it removes routine administration from teachers — grading tests, tracking attendance, preparing materials — and allows them to focus on what is irreplaceable: developing critical thinking, ethics, and human skills.

The human factor remains key

Despite all the potential, a fundamental principle holds: the human must remain in the decision loop (Human in the Loop). Agentic AI can screen a thousand resumes or optimize an energy grid, but final decisions — especially those with cultural, ethical, or legal implications — must remain with people. In the European context, this is further required by the EU AI Act, which from August 2026 introduces obligations for high-risk AI systems, including transparency and human oversight.

What this means for Czech companies

Czech businesses are not lagging behind in AI — according to data from the Czech Statistical Office, approximately 11% of companies with more than 10 employees use some form of artificial intelligence, which is slightly above the EU average. The problem, however, lies in the depth of deployment: most remain at basic automation and pilot projects.

Agentic transformation requires three things:

1. Robust data governance. Companies must start treating data as a strategic asset — ensuring its cleanliness, security, and ethical acquisition. Without quality data, even the best AI model won't function.

2. Collaboration across the ecosystem. Czech universities (CTU, Masaryk University, Charles University) produce top-tier AI talent. Companies must build bridges between academic research and business practice. Examples of a good path include Czech AI Factory in Ostrava or the activities of prg.ai in Prague.

3. Responsible regulation. The government should create an environment that supports innovation while also establishing clear accountability for the outputs of AI systems. The EU AI Act is a framework — now it's about how Czechia implements it in practice.

Agentic transformation is not science fiction. It is a reality that is already changing banking, healthcare, and education from Southeast Asia to Central Europe. The question for Czech companies is no longer whether to deploy agentic AI, but how quickly and responsibly to do it.