Top 10 Startups of the Month: February 2026 Edition
The Companies That Made Technology Useful Again
February is rarely a month for grand declarations. It is too short for that, and a little too impatient. By the time February arrives, markets have usually stopped listening to New Year language and started asking a more irritating question: what, exactly, has changed?
This month, quite a bit had. Not always in the loudest corners, either. There were funding rounds, certainly, and the usual confident talk about scale, platforms and transformation. There always is. But the more interesting movement happened elsewhere — in diagnostics, mobility, infrastructure, disaster response, energy systems, clinical tools. In those awkward, stubborn parts of life where technology is forced to stop behaving like an idea and start behaving like a tool.
That distinction matters more than people admit.
A great many companies can explain their ambition. Fewer can explain their usefulness without slipping into theatre. Fewer still can point to something concrete and say: this is where the system improved; this is where friction fell; this is where a professional gained time, precision or clarity; this is where ordinary people might feel the effect, even if they never learn the company’s name.
That, in the end, was February’s pattern.
Five of the companies below are British. Five are global. Some are young in the traditional venture sense. Some are further along. I am not terribly bothered by that. Age matters less than utility. The better question is whether a company, in a given month, moved a real piece of the world forward. Not in theory. Not in a keynote. In practice.
These ten did.
Top 5 UK Startups
1) Wayve — When Autonomy Begins to Leave the Lab Behind
Autonomous driving has had a peculiar public life. For years it was either overpromised or prematurely mocked. There seemed to be very little room between utopian certainty and weary dismissal. One could be forgiven for losing patience with the whole category.
And yet it persists. More to the point, it matures.
Wayve’s February mattered because the company no longer sounded like a research project asking to be admired for its intelligence. It sounded like a business preparing for deployment. That is a different tone altogether. The major funding round was important, yes, but not because large numbers are inherently meaningful. They are not. We have all seen expensive fantasies before. What gave the moment weight was the implication behind it: this was capital aimed at commercial rollout, real-world vehicle integration, and the kind of next-stage execution that forces an autonomous system to prove itself under conditions far less forgiving than a demo.
That is where things become serious.
If autonomous mobility succeeds — and there is still work to do before anyone sensible declares victory — the benefit will not be abstract. Fewer accidents caused by distraction or fatigue. Greater transport access for people poorly served by existing systems. More resilient urban mobility over time. Roads, of course, are unromantic places. They do not care about elegant slide decks. They care whether a machine behaves well when something unexpected happens.
February suggested that Wayve is moving closer to that harder conversation.
“Technology grows up the moment it must earn trust in the ordinary world, not merely attention in the extraordinary one.”
2) Chemify — Turning Chemistry Into a Faster Conversation
There are industries where delay is accepted far too casually. Drug discovery is one of them. Advanced materials research is another. Everyone likes to talk about innovation; fewer like to dwell on the long, exhausting stretch between an idea and something that can actually be tested, produced or improved.
Chemify is trying to compress that distance.
Its February progress pointed to something more substantial than a typical deep-tech headline. Automated molecular design and synthesis may sound abstract at first glance, but its significance is very human. If useful molecules can be designed, created and refined more quickly, then the downstream effects touch medicines, therapeutics and materials science in ways that are not especially glamorous, but are undeniably important.
This is the sort of company that will never win on spectacle, and perhaps that is to its credit. Chemify is not trying to make chemistry sound magical. It is trying to make it less wasteful. Less dependent on delay. Less trapped by the small inefficiencies that, over time, become very large costs.
There is a quiet intelligence in that approach. Not every breakthrough arrives as a dramatic discovery. Some arrive as a better system for reaching one. February suggested that Chemify is becoming exactly that: a serious attempt to make the scientific pipeline more programmable, more repeatable, and a little less bound to the old rhythm of waiting.
“In science, speed is not vanity. Sometimes it is simply the difference between promise and arrival.”
3) FYLD — Giving Infrastructure the Ability to Notice Itself
One of the stranger habits of modern infrastructure is that we expect vast systems to function intelligently while feeding them poor information from the field. Delays, missed hazards, duplicated work, safety failures — much of it comes down to one very old problem: the people making decisions are often too far from the reality those decisions are meant to govern.
FYLD sits in that gap.
Its work is not about replacing frontline teams with algorithmic fantasies. That is the sort of nonsense one hears from people who have never spent time near a live project. It is about equipping those teams with better awareness, better coordination and better timing in environments where mistakes are physical. Utilities, construction, water, energy — these sectors do not benefit from fashionable language. They benefit from fewer surprises.
That is why FYLD’s February stood out. It pointed to a version of AI that is not performing intelligence for an audience, but applying it where confusion has consequences. Better field awareness can mean safer operations, fewer costly interruptions, more reliable project delivery and less managerial guesswork masquerading as control.
This may not inspire dramatic applause in a conference hall. It does something more useful. It gives infrastructure a kind of nervous system — not perfect, not all-seeing, but far better than the silence many large systems still operate under.
“The most useful intelligence is often the kind that reaches the problem before the meeting about the problem does.”
4) tem — Energy Markets, Minus Some of the Needless Friction
Energy markets have a remarkable ability to make inefficiency sound inevitable. Complexity, opacity and friction are often treated as natural features of the system, as though no one ever designed any of it and all we can do now is endure.
That has always been a convenient fiction.
tem is working on the transaction layer — the part of the market most people never see, yet end up paying for anyway. It is not the most dramatic company in this month’s list, but it may be one of the more practical. If businesses can navigate energy purchasing and management with fewer hidden inefficiencies, the benefit is immediate. Lower friction. Better visibility. Less money lost in systems that became cumbersome simply because everyone got used to them.
I find companies like this quietly reassuring. They are not trying to “reimagine the future” in language borrowed from marketing decks. They are trying to make an existing system less absurd. That is often a far more mature ambition.
Energy, after all, is not an abstract market for most people. It is a recurring pressure. And for businesses already stretched by wages, rent, borrowing costs and uncertainty, even modest improvements in energy efficiency and pricing logic can matter more than a hundred futuristic slogans.
“People begin trusting systems again when those systems stop charging them for being difficult to understand.”
5) Artificial Labs — Insurance, Without the Administrative Fog
Insurance has long enjoyed the protective camouflage of complexity. Processes remain slow, fragmented and stubbornly manual, and much of that inefficiency survives under noble-sounding language about prudence and discipline.
Artificial Labs has chosen to challenge that arrangement.
Its significance in February was not that it suddenly made insurance exciting. No sane person should require that. Its significance was that it continued pushing a deeply manual part of the economy towards clearer, more digital workflows in broking and underwriting. That matters more than it first appears.
Insurance is one of those quiet dependencies modern life takes for granted. Businesses need it to operate. Certain industries are shaped by it. Entire decisions about risk, cost and access move through it. When the infrastructure behind quoting, underwriting and broker workflows becomes faster and more coherent, the effect ripples outward. Friction drops. Delays shrink. More of the process starts to resemble a system rather than a ritual.
This is not a warm story. It is not meant to be. But it is meaningful, particularly in an economy where too many essential sectors still behave as though inefficiency were proof of seriousness.
Artificial Labs stood out in February because it did not promise reinvention in grand language. It simply kept removing old excuses.
“An industry becomes modern not when it talks about data, but when it stops hiding behind delay.”
Global — Five Startups Reshaping the Landscape
1) Aiforia — A Better Diagnosis Is Still One of Technology’s Highest Uses
The technology world has spent so long speaking about transformation that it occasionally forgets to ask a simpler question: what, exactly, is being improved for a human being at the point of need?
Aiforia gives a clean answer.
Its February progress in gastric cancer diagnostics is the kind of development that feels anchored in reality from the start. Not because it is dramatic, but because it is clinically legible. Better support for pathologists means faster, more consistent interpretation in an area where uncertainty is costly and time is rarely generous.
There is an important distinction here. The value of AI in medicine is not that it creates some grand automated future in which human judgement disappears. Quite the opposite. Useful medical AI should sharpen human expertise, not attempt to replace its moral burden. It should help a clinician see more clearly, decide with greater confidence, and reduce the chance that something important is missed in a moment when it should not be.
That is why Aiforia belongs near the top of this month’s list. It reminds us — gently, but firmly — that technology at its best is not a performance of intelligence. It is support for difficult work. And in healthcare, difficult work deserves the best tools available, provided they arrive with discipline rather than arrogance.
“In medicine, the most meaningful innovation is often the one that helps a human notice what must not be overlooked.”
2) THINK Surgical — Precision That Has Finally Entered the Room
Surgical robotics has often been sold with too much performance and too little patience. That is understandable. The machines are visually impressive. The language around them tends to become excitable. But the real question has never been whether the technology looks advanced. It is whether it earns trust in a clinical setting.
That is why February mattered for THINK Surgical.
The move into real clinical use is where this category stops being speculative theatre and becomes something much more serious. Once a system enters live procedures, the conversation changes. Precision is no longer a promise displayed on a stage. It becomes part of a workflow, subject to scrutiny from surgeons, hospitals and, in the end, patients.
That does not mean robotics automatically improves medicine. Medicine is not improved by glamour. It is improved by consistency, reliability and support tools that behave well under pressure. What THINK Surgical represents this month is not the fantasy of a robotic operating theatre taking over from clinicians. It is the quieter possibility that certain forms of precision support may be becoming mature enough to matter in practice.
That, frankly, is more impressive than the spectacle ever was.
“In surgery, progress begins when technology stops trying to impress and starts trying to assist.”
3) ICEYE — Disaster Response With a Better Sense of Time
Floods punish hesitation. They also punish poor visibility. A system that understands impact too slowly will always react as though it is arriving late to its own responsibilities.
ICEYE’s February development stood out for exactly that reason.
Making advanced flood impact intelligence more accessible is not a glamorous frontier in the conventional venture-capital sense. It is, however, a deeply civilised one. Governments, insurers, emergency teams and communities all benefit when the picture of a disaster arrives faster and with greater clarity. Better information does not stop the water. It does help societies respond before confusion becomes its own secondary catastrophe.
There is something admirably straightforward about this use case. It does not pretend to reinvent humanity. It attempts to make a volatile world more navigable at precisely the moments when people are most vulnerable to uncertainty. That is a serious public function, whether or not the language around it sounds fashionable.
In a century that is unlikely to spare us climate disruption, tools like this will matter more than many louder categories. Not because they are more exciting, but because they are more useful when events turn ugly.
“The worth of a system is often revealed by how quickly it can make reality legible in the middle of disorder.”
4) Axelera AI — The Hardware Reminder Beneath the Software Noise
AI discussions still have a tendency to float above the physical world as though energy, heat and hardware were faintly impolite topics. They are not. They are the terms on which the entire industry either scales or embarrasses itself.
That is where Axelera AI enters the picture.
Its importance in February lies in its focus on efficient AI hardware at a moment when the market is discovering, rather belatedly, that not every useful model can run in an expensive cloud stack with heroic energy assumptions behind it. Real deployment means hospitals, transport systems, industrial sites, urban infrastructure and local environments where efficiency is not a technical footnote but the whole viability test.
In that sense, Axelera AI is doing something corrective. It is forcing the conversation back towards physics. Towards cost. Towards the practical conditions that determine whether a technology reaches broad utility or remains an expensive habit reserved for the well-capitalised.
This is not the kind of story that usually dominates a front page. It probably should appear there more often. The software layer receives most of the attention; the hardware layer quietly decides what is sustainable, what is scalable and what was merely overindulgence disguised as progress.
“Every digital revolution arrives, sooner or later, at the same old question: what does it demand from the real world in return?”
5) VulnCheck — Protecting the Systems People Cannot Afford to Lose
Security is one of the least celebrated forms of usefulness because success tends to look like nothing happened. No outage. No compromise. No public embarrassment. No sudden revelation that an essential system was far more fragile than anyone wanted to believe.
VulnCheck operates in that quiet territory.
Its February stood out because it reflects a truth many sectors are still learning: as digital systems become more capable, they also become more exposed. Every new layer of automation, every new connected service, every new intelligent workflow expands the number of things that can go wrong if vulnerabilities are understood first by the wrong people.
That makes exploit intelligence a deeply practical service. Not romantic, certainly. But important. Critical infrastructure, businesses and institutions cannot function safely on optimism alone. They need better visibility into where weaknesses are forming and how risk is evolving.
There is nothing especially theatrical about this. Which is precisely why it deserves attention. The more modern a society becomes, the more it depends on systems remaining intact under pressure. Security work is often invisible right up to the moment it fails. Then suddenly everyone remembers it matters.
VulnCheck earns its place here because it belongs to that less glamorous, absolutely necessary category of progress.
“Modern life depends not only on what our systems can do, but on how well they withstand being tested.”
Cast Your Vote
Which of these startups actually earned spotlight?
Choose the team you’d trust to make a real dent in the world
Vote here: https://www.facebook.com/prime.economist
Editor’s Choice — Aiforia
Not because it was the loudest.
Not because it carried the largest number.
But because February, for all its noise elsewhere, offered a useful reminder: technology is at its best when it improves judgement where judgement matters most.
Aiforia’s work in cancer diagnostics sits squarely in that category. It is not ornamental. It is not an AI parlour trick dressed in medical language. It is a serious attempt to support clinical decision-making in an area where speed, consistency and clarity can make a material difference.
There were bigger rounds this month. There were flashier categories too. But usefulness should outrank spectacle from time to time, if only to keep the market morally awake.
Aiforia did not ask to be admired in February.
It asked to be used well.
“The future will not be built by the noisiest tools, but by the ones people quietly learn to rely on.”
