David Ordonez, Deep Tech Investor at NATO Innovation Fund

What are the most interesting areas of deep tech and why? 

Fundamentally, all areas of deep tech are interesting, particularly those that are solving the greatest challenges we face today! If I had to highlight a few: 

1. Compute: Solving the compute bottleneck is probably one of the most compelling areas right now. The demand for processing power is outstripping supply, and the companies that crack this will unlock enormous value across every other sector. 
2. Robotics & Physical AI: This space is reaching its ChatGPT moment. Falling hardware costs, rapidly maturing AI models, and structural pressures like ageing populations are moving robots from flashy demos into real commercial deployment. The opportunity here is massive. 
3. Manufacturing: We are experiencing a manufacturing renaissance in Europe. Decades of offshoring critical capabilities in an increasingly globalised world made us forget that our ability to produce physical products underpins our society and economic growth. That realisation is now driving a wave of reinvestment and innovation. 
4. Space: The space economy is going through a full-scale industrialisation. As more companies rely on space assets for communications, navigation, and everyday services, the opportunity for startups building that infrastructure grows with it. 
5. Energy: It is clear that we need to transition to a more sustainable world and increase resilience in our energy supply. Novel energy sources like nuclear fusion are attracting extraordinary attention, with entire countries pouring their resources into making it a reality.

What sectors or trends are you most excited about in the next 12-18 months? 

Robotics and Physical AI are probably the most compelling fields in technology right now.  

The macro case is hard to ignore: 

1. Productivity growth has stagnated
2. Populations are ageing across the developed world
3. GDP growth is sluggish
4. We are seeing a decrease in output from physical industries

These are structural problems, and they all point in the same direction: we need machines that can do more in the physical world. 

What makes this moment different is that we're approaching what I'd call the ‘ChatGPT moment’ for robotics. Compute is getting cheaper and faster, and the cost of components like sensors and actuators is dropping consistently. The models underpinning physical intelligence: vision language-action systems, simulation-based training, real-time spatial reasoning, are maturing very rapidly. 

Over the next 12–18 months, I expect we'll see both generalist and specialised robots deployed at meaningful scale across fulfilment centres, manufacturing lines, and public roads. Critically, this won't just be a story about large enterprises. Small businesses will be able to automate tasks and processes that were previously only viable at a large scale. 

What excites me most is the sheer breadth of the opportunity. Most companies in this space are still building everything from scratch - data pipelines, navigation, manipulation, simulation. Every layer of that stack is an opportunity and the window to build foundational companies is wide open. 

Has the rapid rise of generative AI impacted your investment strategy? 

AI has fundamentally changed how capital should be allocated. Companies that used to require years of development and large teams of specialised talent can now compress those timelines dramatically.  

This reframes what deep tech investing actually looks like. The old label of "patient capital" - implying slow, uncertain, and drawn out - is outdated. Deep tech is now attracting long-term capital with the ambition to move fast. Essentially, AI compresses timelines without compressing the depth of what's being built and that changes the return calculus, which in turn benefits investors, LPs, entrepreneurs, and the whole ecosystem. 

Are there any under-the-radar technologies or sectors you think are poised for rapid growth?  

Resilient PNT and navigation alternatives: GPS underpins virtually every critical system in  modern life - logistics, financial transaction timing, telecommunications, defence - and yet it  remains remarkably fragile. Signals are weak, easily jammed, and increasingly targeted by adversaries (Ukraine as an example). A new generation of companies building positioning, navigation, and timing (PNT) solutions that don't depend on satellite constellations at all, from chip-scale atomic clocks to quantum sensing approaches. The defence and intelligence communities have been funding early work here for years, but the commercial pull is accelerating fast - from autonomous vehicles that can't afford to lose positioning, to critical infrastructure operators under new regulatory pressure to demonstrate GPS-independent resilience. 

Power electronics: This sector rarely makes headlines, but it sits at the bottleneck of nearly every major energy transition trend. The core problem is simple: electrical energy almost never arrives in the form you need it. It has to be converted, and every conversion stage loses energy as heat. Take AI datacenters for example, where even small efficiency losses translate into hundreds of megawatts wasted and hundreds of millions in added cost. The same problem plays out across EVs, solar, fast charging, and grid storage. New silicon compounds like silicon carbide and gallium nitride are unlocking step-function improvements in efficiency and power density, and the companies that control these critical nodes in power conversion are positioned to command outsized value as electrification scales. 

How do you think the current macroeconomic environment is shaping the deep tech investment space? 

The current macro backdrop is, paradoxically, both a headwind and a tailwind for deep tech. 

On the headwind side, higher interest rates have made it harder for capital-intensive deep tech companies to raise on vision alone. The era of cheap money that powered speculative bets on moonshots is behind us, at least for now. That means founders need to show clearer paths to revenue, and investors are applying more rigorous scrutiny to technical milestones and commercial traction before writing large checks. 

But the tailwinds are arguably more powerful. First, geopolitical fragmentation is creating  enormous demand for supply chain sovereignty - in semiconductors, critical minerals, energy systems, and advanced manufacturing. Governments are deploying historic levels of industrial policy capital through mechanisms like the CHIPS Act, the Inflation Reduction Act, and the EU Chips Act. For deep tech companies aligned with these priorities, there's a depth of non-dilutive funding and guaranteed demand that simply didn't exist five years ago. 

Second, the AI infrastructure buildout is pulling forward investment timelines across the entire deep tech stack - creating downstream opportunities in energy, cooling, advanced packaging, optical interconnects, and novel memory architectures that move much faster to market than they would have in 2020. 

Third, and perhaps most importantly, the macro environment is acting as a natural filter. In a  zero-interest-rate world, deep tech competed for capital against software companies showing hockey-stick growth with minimal capex. Today, investors are relearning that durable competitive moats often come from hard-to-replicate technical differentiation rather than network effects or growth hacking. That reappraisal favours teams solving genuinely hard problems where the intellectual property itself is the barrier to entry.

Do you have any advice for founders on how to get noticed in the deep tech space? 

Deep tech founders, especially in Europe, are chronically underselling themselves. We have  world-class science and genuinely breakthrough technology, but too many founders treat  storytelling as beneath them. That needs to change. Deep tech is hard to sell by nature because you're often asking people to believe in something that won't fully exist for years. Your customers can't touch it yet. Your investors can't see the revenue yet. The only thing bridging that gap is your ability to articulate where this is going and why it matters. 

Look at the founders who consistently attract the best talent, the best partners, and the best capital. They're not just technically brilliant - they know how to craft a narrative that pulls people into their world. There's a fine line between storytelling and overselling, and the best founders learn to walk it. They sell the vision, build a credible roadmap, and then execute relentlessly against it. Every milestone hit makes the next chapter of the story more believable. 

If you're sitting in a lab with a genuine breakthrough and you can't communicate its value, it won't get noticed. Technology that goes unnoticed achieves no impact in society. 

What mistakes do companies make when communicating to investors?

I see three mistakes come up again and again. 

1. Being unclear about the problem you're solving and why it matters. You'd be surprised how many pitches I sit through where a founder spends twenty minutes on the technology and thirty seconds on the problem. Investors aren't buying your technology, they're buying into the change it creates. If I can't explain in one sentence what problem you solve and why anyone should care, we have an issue.  
2. Having no long-term vision. Too many founders show up with a plan that covers the next twelve to eighteen months. When investors back a company, they are entering a partnership that often lasts longer than the average marriage. So if your pitch only extends to your next funding round, it’s a problem. The best founders have a clear picture of the world they're building toward over a five, ten, even fifteen-year horizon.  
3. Having no credible roadmap connecting today's reality to that long-term vision. Investors need to see the steps in between. How do you overcome the short-term realities? How do you generate revenue today while building toward something that might not fully mature for years? The founders who get this right show you the bridge: here's where we are, here's the near-term milestones that prove the technology works, here's how we start generating revenue early, and here's how each step compounds toward the ultimate goal.