Mateusz never expected his career to follow a traditional path.
Growing up in Poland, he attended one of the country's oldest universities. While respected academically, it wasn't considered a traditional pipeline into elite consulting firms. Breaking into McKinsey from a non-target school would require persistence, patience, and a willingness to prove himself repeatedly.
"It took a lot of time," he says. "Coming from a non-target background meant I had to work harder to get noticed."
That’s still an accomplishment he's most proud of. Not because of the brand on his resume, but because it was proof that he could compete at that level despite starting from a less conventional position.
Today, Mateusz works across AI-focused roles with Mercor and Deloitte's AI & Data practice, combining consulting experience with frontier technology. The common thread throughout his career has been curiosity.
"I don't like being bored," he says. "I need to learn something new or solve a new problem every now and then."
For someone motivated by constant learning, AI has proven to be the perfect environment.
The human side of building AI
Many people assume that AI work is primarily technical, focused on engineering, coding, or evaluating model outputs against predefined criteria.
Mateusz sees it differently.
"The most important thing we contribute isn't just domain expertise," he explains. "It's the human feeling."
When reviewing AI outputs, he isn't simply checking for correctness. He's asking whether it reads like a real, natural conversation. The goal is to create answers that feel like it came from a professional who just knows.
That perspective comes directly from years of working with clients, where the stated question is rarely the whole question.
"Very often, behind a simple ask there is much more depending on the broader context," he says. "Understanding what someone actually means—not just what they wrote—is increasingly important."
Learning what AI can and can't do
Working directly with frontier AI systems has also changed how Mateusz thinks about the technology itself.
Before Mercor, he saw the technology the way most people do. Impressive, but somewhat abstract. Now, after spending countless hours evaluating and improving model behavior, his perspective has become more nuanced. Not only is he more confident using AI in daily life but he understands where the technology excels and where caution is still necessary.
"I understand where AI works best, what the limits are, where to trust it, and where to be careful," he says.
That exposure has made him more optimistic about AI's future, not less.
"Business is people," Mateusz says. "AI can help us, but connecting with others, understanding tradeoffs, and making decisions under uncertainty are still fundamentally human skills."
That's why, for him, the most important ingredient in better AI is still human judgment.