Preparing Developers for AI and the future of work

As Onja's Professionalism and Communication Mentor, Adrià Trepat’s job is to make sure our world-class developers remain highly desirable for top employers. Originally from Barcelona, Adrià came to Onja with a background in international business and leadership development. He believes that the right support, given at the right moment, can change the trajectory of someone's career.

As AI transforms ways of working across industries, he’s expanded Onja’s Tech Talk series to address trends in the tech industry and general job market demands. From immediate job-specific technical skills to attributes that foster long-term professional growth, Onja remains committed to nurturing the next wave of tech talent.

Read on to learn more about Onja’s Tech Talks and how we’re helping underprivileged Malagasy youth continue to thrive in their roles as world-class software developers.

1. Tell us about Onja’s Tech Talks. What purpose do they serve?

Tech Talks are something of an institution at Onja. Right now, these are mainly peer-led sessions where developers present on a topic to fellow developers and non-developer staff. We’ve been running the Tech Talks since our first wave of graduates entered the workforce and they've become a much-loved tradition.

One of our developers Rinon Tendrinomena delivered our most recent Tech Talk. He walked the team through the evolution of React rendering from basic client-side rendering right through to React Server Components and how modern frameworks like Next.js let you combine multiple strategies in the same application. And that's the point; these are deep, practical, the kind of thing that rarely comes up when you're heads-down on a project.

Developers also get to practice public speaking and hone their skills in explaining complex concepts clearly, which matters just as much as the tech stacks they’ve mastered. It keeps the team connected, too. Our developers work for different companies, on different projects, but this is a moment to come back to the team: learning from each other and pushing each other forward.

In 2026, we're also bringing in more external speakers to address today’s hot topic-AI.

2. On the topic of AI, what skills are essential for developers today?

Adaptability comes first, and it always has at Onja. Our developers, by the nature of what they've been through with our training programme, excel at adapting to new conditions fast.

Fostering that resilience is something we intentionally build into the Onja experience because in their industry, things that were standard practice six months ago can become outdated. You have to move with it and our Developers have no problem with that.

The second thing is critical thinking. AI is producing a lot of output, and developers need to be able to assess whether what they're getting is actually the right answer to the right problem. Currently, no tool is going to make that call for you. IT comes from the developer.

Then, there's what I'd call big picture thinking, or systems thinking. The job is increasingly less about writing every single line of code yourself and more about understanding the product you're building, the system you're designing, and the problem you're genuinely trying to solve. That high-level thinking is worth developing.

3. How does Onja approach AI in practice and how do developers know when to trust the code it produces?

We block time for developers to explore AI tools together. We'll run sessions to walk through what's new and discuss the possibilities and implications of these tools. We're just as careful about the other side of it: the risks, the limitations, and things to watch out for. It's easy to focus only on what a tool can do and overlook what can go wrong.

Security is one of the big ones. We always remind developers to use tools that have been approved by the companies they work for. There are real safety and compliance considerations, and that's where things can go wrong.

Beyond that, we encourage them to explore freely and to share what they're learning with each other. A lot of the best knowledge-sharing at Onja happens between each other, and AI is no different.

Our developers are already using AI to write code directly, and that's a trend that will only continue as this tech evolves. What we emphasise is this: Don't just accept the output. Understand it. Is it actually solving the problem? Is it creating issues somewhere else in the codebase? Is it adding technical debt that someone's going to have to deal with? There are different ways to check, from reviewing diffs, running tests, or looking at how new code sits against the wider architecture. The right method depends on the situation. But whatever the method, check it.

4. What advice would you give to someone starting to learn coding today?

Focus on learning the fundamentals of computer science. The way we write code has changed, and it'll keep changing. In the last century, people were punching cards and that was programming. Then it was keyboards, writing code line by line. Now, increasingly, developers are directing AI and reviewing what comes back. It's a big transition, but it's not the first one this industry has been through.

What stays constant through all of those transitions is the fundamentals. If you understand how good software is built, how systems are designed, how quality is maintained, how to properly define a problem before you try to solve it, you're going to be in a strong position regardless of what the tools look like. Knowing one specific programming language just isn't the competitive advantage it once was. What counts now is understanding what you're building and why, and to use the tools available, AI included, to get there well. Know the fundamentals. It's more important now than ever.