My name is Hugo and I am the cofounder and CEO of Encube, an AI-powered engineering platform that helps modern hardware teams bridge the gap between design intent and manufactured reality.
I have always been fascinated by the intersection where software meets hardware, and how software can transform the way physical products are imagined and invented to drive meaningful real-world impact. That fascination has driven everything I have done. From autonomous driving research at Robert Bosch in Silicon Valley, to building and scaling industrial software as VP at Sandvik Group, one of the world's leading industrial engineering companies, and SVP at Aker Group, a major European integrated energy company. After almost a decade working at the heart of the hardware industry, it became clear to me that the software powering hardware R&D was built for a world that no longer exists. That was the start of Encube.
The biggest challenge I'm seeing right now is that hardware development is being reshaped by geopolitical realignment, a deepening talent crisis and the accelerating transition toward sustainable production. These structural shifts are reshaping how industrial economies build, produce and compete.
Across Europe, geopolitical tensions have upended long-standing assumptions about security and trade. As global supply chains fracture, countries are racing to rebuild manufacturing capabilities they once outsourced in an effort to secure industrial autonomy. But wanting to bring production home and actually being able to are two different things. What Europe lost when outsourcing production wasn't just production capacity, it is the know-how that made production possible in the first place. The engineers who carry that knowledge are retiring faster than we're able to replace them.
The second shift is the accelerating talent gap. This is the most under appreciated challenge of the three in my view. Traditionally, manufacturing knowledge has largely lived in the heads of the people doing the work. When a senior engineer walks out the door, so does their cumulative experience and knowledge of their trade. Across Europe, we haven't invested nearly enough in building the next generation of hardware engineering talent. As a result, one third of open job positions are never filled and 7 out of 10 people under the age of 30 can't even imagine a career in hardware. That should worry everyone from individual decision-makers at industrial firms to political policymakers.
Adding to industry pressures, tightening European sustainability regulations are requiring manufacturers to rethink how products are designed and built. Making the right choices requires a depth of manufacturing understanding that most teams simply don't have readily available.
The third shift compounds the first two. Tightening European sustainability regulations mean hardware teams must rethink how products are designed and built. Making the right choices requires a depth of hardware development understanding that most teams simply don't have readily available.
Geopolitical realignment, a shrinking talent base and rising sustainability requirements are converging at the same time. In that environment, the tools hardware teams rely on to develop products can no longer be an afterthought. Because the software that exists today was built for a world that no longer exists.
If we want to secure and strengthen the industrial base Europe once built, we need to give hardware teams the right tools to work with and we need to act now. Because the software that exist today was built for a world that no longer exists.
Reinvigorating Europe's industrial base means inheriting the complexity that comes with it. Hardware development has always demanded that many disciplines work in concert, but the tools teams rely on were never built with that in mind. The consequence shows up in how teams actually work day to day. Feedback is scattered across email chains and messy Microsoft Teams threads, while design reviews are built around static screenshots and PowerPoint slides. Most people involved in the process simply don't have the tools to engage with the actual geometry. As a result, critical context gets lost and by the time the right people are aligned, decisions have already been made that are hard to undo. Businesses are then left with a difficult choice to either accept higher production costs than the business case ever accounted for or redo the designs and delay market launch.
After spending almost a decade building the software division at Sandvik Group and later leading a similar transformation at Aker Group, I encountered the same challenges across the organisations I met. From the world's largest industrial engineering organisations to a small job shop in Ohio with five guys and a dog. Decisions are made in isolation because the existing workflows makes alignment genuinely hard. Legacy systems aren't the problem in themselves, the problem is that they weren't designed to surface trade-offs across the full value chain at the point when they actually matter, which is at design. What's left is a fragmented toolchain and feedback loops that are far too slow for the pace modern hardware development demands.
Encube brings everything together on a single canvas, making it easy for hardware teams and everyone that needs to be involved to have the right context at the right time. Not by replacing the tools teams already rely on but by connecting them around the decisions that matter most. Our AI surfaces manufacturability risks, cost drivers and trade-offs directly at the point of design in a form factor every team member can engage with. This means that the right people are always in the same conversation, with the information they need to make more informed decisions faster.
The main challenge I have with the use of expressions like digital thread is that they sound great at first glance and on paper. But, the devil is in the details, as the saying goes. What do we actually mean by digital thread? If you ask a mechanical design engineer, a manufacturing engineer and a quality inspection engineer, I'm not convinced you would get the same answer. And therein lies a problem for the industry. We're very good at coming up with high-level abstractions like digital thread, industry 4.0, Internet of things, digital twin and so on, but we often forget what it is that we want to accomplish when we embark on these transformation projects.
If we consider the journey from concept to manufacturing-ready set of assets, the digital thread, from my perspective, should articulate a company's desire to ensure that engineering intent remains traceable and intact as it passes through every function that touches a product. From design, through process planning, manufacturing, inspection, and back.
Consider something every hardware engineer has lived through: the Engineering Change Order, ECO. A design engineer releases an ECO that changes a critical dimension on a component. That change touches the drawing, the process plan, the CNC program, the inspection plan, the fixture design, maybe a supplier spec. The question the digital thread should answer is: can you, right now, tell me every downstream artifact that was affected by that change, and confirm that each one has been updated?
In most companies, I believe the honest answer is no. The ECO gets released in the PLM system. Someone sends an email to manufacturing engineering. Someone walks a printed markup to the shop floor. The inspection department maybe gets CC'd. And six weeks later, parts fail inspection because the CMM program was never updated, or the supplier is still building to the old spec because nobody remembered to send them the new revision. That's not a technology failure. It's a traceability failure. The intent changed at the source, while the thread connecting that source to everything downstream was held together by emails, tribal knowledge, and good intentions.
So when someone asks me about the digital thread, I think the real questions are: when engineering intent changes at the source, can you identify everything downstream that needs to change with it? And can you confirm that it actually did? That's the digital thread, stripped of all the buzzword baggage. It's not about any specific technology or methodology. It's about whether your organization can answer those two questions reliably, or if you're relying on Joe remembering to tell Bob.
The reason this is hard is ultimately organizational, not due to technology. The thread breaks at the boundaries between functions because those functions typically own different systems, follow different processes, and often report to different parts of the org chart. The design engineer works in one world, the manufacturing engineer in another, and the connective tissue between them is often a PDF and a prayer. We need more intuitive software tools to enable this. That’s why Encube exists.
The risk doesn't announce itself. It hides in the small, seemingly innocuous decisions made early in the process. When it finally surfaces, it's because something broke or because
someone caught it late enough that fixing it means going through a long and costly loop. Either way, hardware teams are left with the same difficult choice. Absorb production costs that were never part of the business case or go back to the drawing board and put the market launch at risk.
I have two concrete examples from my time at Sandvik and Aker. The first was during a site visit to one of the world's largest truck manufacturers. On the shop floor, people in production shared a nightmare scenario with me where a single hole diameter was 0.15mm too large, but that hole was repeated 400 times on each part. Somehow, this had slipped through unnoticed. The misalignment ended up costing the company millions of dollars compared to the original business case. All because of one small, overlooked detail in the early design phase.
The second was at Aker, where we were running large-scale offshore projects in Northern Norway. Two critical mating components, roughly half a meter in total size, inside a 50x50 meter rig, had an alignment issue of about two centimeters. Small enough to miss in a design review and large enough to make on-site assembly impossible. The consequence wasn't just fixing those two parts, it meant redesigning the entire section and we lost a full year to rework.
These aren't edge cases. They're what happens when design, cost and manufacturing operate without a shared view of the product. In highly regulated and mission-critical industries the stakes make that disconnect even harder to absorb. By the time the problem surfaces, the cost of fixing it has already grown far beyond what anyone planned for.
When we talk about fixing the broken digital thread it can sound abstract, but in practice it is quite concrete.
Encube gives hardware teams a shared environment where mechanical engineers, manufacturing leads, cost engineers and suppliers all work from the same underlying data. Design intent, manufacturability analysis and cost implications are visible to everyone at the same time, so discussions move from what version are we looking at to which is the right decision. An engineer reviewing a design change can immediately see the downstream impact on manufacturability and cost without waiting weeks for a separate review cycle.
A big part of what that unlocks is how engineers actually spend their time. Today, too much time goes toward manual tasks that are tedious and error-prone. Hunting for differences between design revisions and chasing missing information that should already be available. Our agentic AI does the heavy lifting, so engineers can focus on inventing better products, make smarter design decisions and move faster in an environment where the pressure on price and time to market has never been higher. A major Swedish truck company we partner with built a business case for Encube where our AI reduced manual design review and manufacturing handover by more than 70% within a month of rollout.
This shift matters even more when you consider what the talent gap looks like up close. One of our customers recently told me they don't know if they'll be able to keep their factory running. Not because they lack work, they have more orders than ever, but because their most experienced engineers have retired or are about to and they simply can't find people to replace them. When knowledge is embedded in the workflow, rather than sitting in someone's head, it doesn't walk out the door when people leave. That's what keeps me going and why I think the timing for what we're building is now.
Most generative AI tools are very good at language but poor at grounded reasoning. They learn patterns in data and predict what comes next based on statistical likelihood. That works well in many contexts but it runs into a fundamental problem in hardware engineering where the hardest challenges require grounded reasoning about the physical world.
In software, you can push an update if something breaks but in hardware you can't. The products hardware teams build end up in places where errors carry far more risk than they typically do in software. A brake system that fails isn't a bug you patch over the air. That changes everything about what you can ask AI to do in hardware development and how much you can trust its outputs.
The core of what we've built is AI that reasons from physics and manufacturing constraints and that's a fundamentally different foundation. It's what allows our AI to reason about designs, surface real trade-offs between form, fit and function and produce outputs that teams can act on with confidence. In software development, the feedback loop is tight because you have a compiler, automated tests and an execution environment that tells you whether something actually works. Hardware has never had that equivalent and in its absence we've had to build it. A compiler for hardware. One that tells you definitively whether a design is manufacturable and at what cost rather than making a statistically likely guess. That's what makes our approach different and it's why hardware teams can depend on what we build.
2025 was a big year for us and the momentum has carried into 2026. We came out of stealth with $23 million in funding in at the end of Q4, backed by industry focused investors such as Kinnevik, Inventure and Promus Ventures. This capital injection allows us to accelerate commercially, speed up product development further and deepen our research in applied AI.
Since January 2026, we've quadrupled the speed of activating new customers on the platform and cut the time from lead to live by 50%. Enterprise logos make up more than 90% of our funnel, which signals that we're solving a problem that matters at scale, and not just for early adopters willing to take a bet on new technology.
On the product side, the most significant step forward has been in agentic AI. We released state of the art research performance on manufacturing aware agentic hardware design and introduced the Hardware Canvas, an entirely new way of engaging with 3D and 2D design assets natively powered by AI. At the heart of it is a growing suite of AI apps, each purpose built to solve a specific challenge hardware teams face today, from identifying how design changes affect the balance between form, fit and function to flagging manufacturing risk and excess cost before designs are frozen. Together these represent a meaningful shift in what's actually possible. AI that doesn't just assist but autonomously analyzes designs across revisions and generates clear actionable next steps the entire team can act on.
What I'm most encouraged by is the nature of how customers are using the product. Teams aren't treating Encube as a periodic check or a handoff tool. They're using it as a shared decision surface in their day to day work, which is exactly the behavior we set out to create from the beginning.
The threat people should actually be worried about is not automation. It is that the manufacturing know-how the industry depends on is retiring faster than we can replace it. Years of outsourcing combined with underinvestment in the next generation of engineers is where that problem starts. That knowledge lives in people's heads, not in systems. When those people leave, it leaves with them.
For many companies this is a question of both survival and competitiveness. It is worth being clear that demand for hardware has not gone down. Every company I talk to has more work than they can handle. The challenge is not a lack of orders, there is a lack of people to do the work. If you can not capture and transfer that knowledge and make it available to everyone before it walks out the door, your ability to keep up goes with it.
The second part is what happens to the engineers who remain. Too much of their time goes into manual and time-consuming work that drives up costs and slows things down. It is also work that has probably very little to do with why they became engineers in the first place. If AI can take that off their plate, engineers can spend their time on the problems they are actually passionate about and where their judgment makes a real difference. That is what AI should solve. To me, this is empowerment, not replacement. In an industry already struggling to find the people it needs, it is a necessity.
The next three to five years will force a reckoning that has been building for a long time. Europe is attempting to rebuild an industrial base it spent decades hollowing out and the gap between ambition and execution is significant. Wanting to reindustrialise and actually being able to are two very different things.
The talent crisis will get tougher before it gets easier and the complexity of what needs to be built is only increasing while release cycles are getting shorter. We will not be able to teach and graduate our way out of it. The answer has to be a fundamentally new generation of software built for the world as it actually exists today. One that brings real manufacturing intelligence and agentic AI into the design process in a way that hardware teams can actually depend on. I don't think that is the only part of the answer, but it is an important one.
There is an enormous amount of hardware that still needs to be invented and for a long time the tools simply haven't been good enough to keep up with the pace modern hardware development demands. That is changing. Just as in previous technology shifts, when the early majority starts to adopt a new paradigm it becomes very hard for those who wait to close the gap. The companies that move now will be the ones that define what hardware development looks like for the next decade. That is the reality hardware teams are facing today and it is why we built Encube.
The single most important thing engineering and manufacturing leaders should be paying attention to right now is the gap between the pace at which their industry is changing and the pace at which their tools and processes are keeping up.
As I have touched on earlier, the market structure is changing. Geopolitical realignment is forcing companies to rebuild supply chains and manufacturing capabilities they spent decades outsourcing. The talent carrying the deepest manufacturing knowledge is retiring faster than it can be replaced. Sustainability requirements are tightening and at the same time the complexity of what needs to be built is increasing. In this environment, the ability to control cost and get to market fastest are the two factors that determine whether a company stays competitive or falls behind.
The leaders who will come out ahead are the ones who treat this as a systems problem rather than a series of individual challenges to manage. In practice that means three things. First, investing in tools that give teams a shared view of trade-offs early in the design process. Second, finding ways to capture and transfer manufacturing knowledge before it retires. Third, taking AI seriously as a practical lever for closing the gap between what teams know and what they need to know to make better decisions faster.
The companies that figure this out now will be in a fundamentally stronger position in five years. The ones that don't will find it very hard to close that gap from behind.
Hugo Nordell has always been fascinated by the intersection where software meets hardware and how this connection can transform the way physical products are built while driving real-world impact. That belief led him to study physics, industrial engineering and economics and later took him from frontier tech in Silicon Valley to connecting active volcanoes to the internet and back to Europe to pioneer the future of how hardware gets built.
In Silicon Valley, Hugo worked on everything from autonomous vehicles and drones to volcanic early-warning systems designed to protect millions of people living near active volcanoes. His work soon caught the attention of the board of Sandvik Group, one of the world’s leading industrial engineering organisations, and he was headhunted to Europe to lead its digital transformation. During his tenure, Sandvik’s share price more than doubled and he built software development teams across eight countries, reaching customers in over 100 markets.
That success led to his recruitment into a new executive role at Aker Group, a major industrial holding company in the Nordics, where he was brought in to drive a similar transformation. Across both roles, Hugo saw the same pattern: up to 80% of a product’s cost is set once the design is locked and even small missteps can cause costly problems later in production, forcing companies to choose between lost profit or delayed launches.
This marked the beginning of a larger mission. Driven by a bold vision to rethink how products are developed and help hardware teams face the demands of a rapidly changing world, Hugo left the corporate world to build Encube together with Johnny Bigert.
