*By Filippo Di Cesare
You invest in AI, hire modern software, and build some prototypes. Three months later, you discover the results aren't up to par. Sound familiar? It's not that the technology failed. It's that you may have trained the AI for your business, but you haven't yet trained your business for the AI.
- Training AI for Business It's based on clear use cases: reducing costs, accelerating processes, improving service. It works, but it's limited to the "here and now."
- Train the business for AI It's a different ballgame: it involves preparing the company's culture, data, processes, and even value model so that AI not only addresses current pain points but also allows for reinventing products, services, and even the way we compete in the market. It's about accepting that the customer's problem can change, and that the true differentiator will be the ability to adapt.
In Brazil, investments in artificial intelligence (AI) are expected to exceed US$1 billion by 2026, according to the International Data Corporation (IDC). A recent Gartner study reveals that 641,000 technology executives worldwide plan to implement agentic AI in the next two years. In Brazil, however, few projects along these lines have actually begun. Even so, the same study indicates that more than 681,000 Brazilian companies intend to develop agentic AI initiatives in the same period, exceeding projections for other countries.
But the central question is not how much to invest, it is how is invested in. Are we simply feeding algorithms and automating tasks, or are we preparing organizations to absorb and multiply the impact of AI?
It's necessary to look at all aspects to prepare a company for this new development, and this training involves several dimensions:
- Data as a strategic asset (governance and reliability first and foremost);
- Open integration (APIs, interoperability, avoiding silos);
- Continuous learning mindset (fail fast, adjust fast);
- Ethics and governance (AI as a vector of trust, not risk);
- Talent and culture (professionals who think with AI, not just about it).
And where to start?
The first step isn't buying technology. It's mapping out a strategic and relevant problem where the impact of AI can be clearly perceived, while simultaneously preparing data and people around that problem. Small enough to learn quickly, but large enough to demonstrate value.
This is the “entry point”: a use case with real value that, beyond the immediate result, helps create the culture, data, and learnings to scale AI within the organization.
The question then remains: do you want to simply train models for today's problems, or do you want to prepare your company for challenges that don't even exist yet? Because, in the end, AI won't just answer the business, it will redefine it.
*Filippo Di Cesare is the group's LATAM CEO Engineering, a global Information Technology and Consulting company specializing in Digital Transformation. With a degree in Economics and Statistics from the University of Bologna in Italy, the executive has worked for over two decades in digital strategy and operations and has led projects at major market players such as TIM, Claro, Sabesp, Eletrobras, Nestlé, Volvo, and Pfizer, among others.
Notice: The opinion presented in this article is the responsibility of its author and not of ABES - Brazilian Association of Software Companies
EN 
PT 
