Most defence AI is validated once and expected to stay sufficient after that. In the field, conditions change too fast for that to hold. NestAI is building NestOS, an open, modular platform for adaptive intelligence in European defence. AI that learns from the field and evolves after deployment.

"You need to be comfortable with ambiguity. The problems we solve don’t come with textbook answers. You’re figuring things out in a field where no one has done this before."

That's Chinh Nguyen describing the work he does at NestAI. A year ago, he joined as employee number 35. The company has since grown to over 200 people. In addition to the company evolving and growing, Chinh's work as an engineer at NestAI has turned out to require a mindset of constant learning and adaptation. This is something he finds the most exciting part of the work.

Originally from Vietnam, Chinh moved to Finland in 2016 to study before building a career in AI and computer vision. He was happy at his position, with a clear focus and forward path.

Then the phone rang.

"What made me decide to join was a call from the founder, Peter Sarlin. He had just started this new initiative to strengthen European defence. I just couldn't say no, even though I'd just told my previous employer that I'd probably be there for the next five or ten years, and I loved being there. We’re in a moment in time when we need to invest in defence and AI capabilities. So contributing to that mission as part of NestAI felt meaningful."

It's the mission and the interesting technical challenges that have kept Chinh at NestAI.

‍"We develop technology that has to keep evolving, because the adversary evolves too. It's a constant cycle of adaptation and counter-adaptation. You never run out of things to do. We build detection systems, but the adversaries try to disrupt them, camouflage against them, spoof them. So it's a constant race. There's nothing that's going to be set in stone and used forever. And that's what keeps things interesting and motivating."

‍

Building around real-world constraints

Defence environments create constraints that commercial software rarely has to deal with. There's no copy-paste when operating in the context of constantly changing conditions. This is where the understanding of the real world constraints and deep technical expertise merge.

"An interesting technical problem is selecting the right spectrum for computer vision solutions. We have different wavelengths - short, medium and long - each with unique strengths and weaknesses. So it's about how to fuse those sensor signals. But selecting the right combination is not always easy, because you have to work within compute constraints and data limitations.“

The physical world brings its own constraints, too: GPS isn't always available, and weather and lighting can vary dramatically. On top of that, operational contexts create different kinds of engineering problems.

“On modern battlefields, emission control is a critical concern. We need to limit what we emit. This means we cannot use wireless communication or use radars, LiDARs or IR illumination for detection in certain situations. Operational constraints thus dictate the technology selections and approaches.”

At one point, the team needed a fast way to verify an algorithm against standard-sized IR detections. Instead of outside-the-box thinking, Chinh and his team turned to inside-the-basketball -one.

"One idea was to fill up a basketball with hot water so it would be entirely visible in IR, and then hold the basketball over your head at various distances to gather test data.”

It’s all about adaptability and creative problem-solving

Ask what skillset the work needs, and the answer covers a lot of ground. At the heart of it, though, is lateral thinking and problem solving.

"You need to be adaptable: one day you're helping with backend, the next an algorithm, or embedded programming. You also need genuine curiosity about the domain, really wanting to understand why things are done a certain way. It's about asking the right questions."

That's where creative problem-solving happens.

Engineering for the real world

Drawing on experience from robotics, AI and more commercial fields, Chinh has his own take on what makes working in defence different.

"Because we're in defence and we do physical things, there are physical limitations. So you need reality checks. Your ideas aren't just about what the hardware can do in a computer; the physical constraints come from the operating environment. So you need strong technical ownership and to be creative, but also be grounded in those constraints."

It’s the same idea with the basketball, really. An idea is only as good as its test in the actual field.

Work with us

We build adaptive intelligence for Europe's security.

Interested in joining us? Check out our open positions here.