The Zerobionic Story: A small Nairobi assistive-robotics startup is building AI hardware
There is a classroom in Garissa County, in northeastern Kenya, where, for the first time in any of the deaf students’ lives, the teacher’s voice is reaching them in real time.
Not through a human interpreter. Not through a transcript on a screen later that evening. Through a pair of 3D-printed robotic hands, mounted on a stand at the front of the room, made largely from recycled plastic waste, listening to a hearing teacher and rendering Kenyan Sign Language to a child who, until this year, had never seen a STEM lesson signed in real time at all.
When the student signs back, a camera reads her hands. The robot speaks her answer out loud for the teacher to hear. A two-way conversation, in two languages, mediated by a machine.
The company that built it is called Zerobionic. It was founded, in 2023, by a 19-year-old undergraduate from Nakuru, in a lab at Strathmore University in Nairobi. He has never taken venture capital. And by April of this year, he had beaten 1,200 applicants from 45 African countries to become Qualcomm’s only Kenyan pick on the continent.
This is the story of Zerobionic.
The Nakuru Kid Who Went to Ghana
To understand Zerobionic, you have to start in Nakuru.
Maxwell Opondo grew up in a small household in Nakuru County — his mother, his grandmother, two younger siblings, the four of them held together by his mother’s work and his grandmother’s example. Money was tight. Opportunity, for a Kenyan kid in a small Rift Valley town, was something you had to manufacture for yourself.
So he did.
At sixteen, while still in secondary school, Maxwell built two things that should not have come out of a Kenyan high school. The first was a project he called JaceBot — an AI-powered system to trace water hyacinth, the invasive aquatic weed that has been suffocating Lake Victoria and a dozen other African lakes for decades. The second was a machine-learning anti-theft system for his own school, which by his own account reduced theft incidents on campus by about ten percent.
JaceBot won him a national science prize. A trip to Dublin. And, more decisively, a scholarship to Strathmore University, in Nairobi, where he enrolled in 2022 to read Informatics and Computer Science.
Strathmore is one of Kenya’s most prestigious private universities. It is also a place that takes its engineering students out of the lecture hall and into the field. And in Maxwell’s case, the field was not in Kenya at all.
Strathmore is connected to a network of prosthetics outreach programmes that send students into hard places. Maxwell joined two — one to Sierra Leone, one to Ghana — fitting amputees with prosthetic limbs. He was a teenager from Nakuru, helping a stranger in Freetown stand up on a leg that had not been there an hour earlier. It is the kind of week that recalibrates a person.
But the moment that mattered was not the fittings. It was what he noticed in the periphery.
In clinics and in classrooms and in the conversations around them, Maxwell kept seeing the same thing happen, again and again. People with hearing impairments were being systematically excluded — not from medicine, not from food, but from opportunity. From higher education. From STEM tracks. From the rooms where futures were decided. And not, he realised, because of any cognitive limit. Because of an infrastructure limit. There were not enough sign-language interpreters. There was no signed vocabulary for advanced science. And there was no assistive hardware that bridged the gap when humans were not in the room.
The way Maxwell himself frames the moment, in interviews afterwards, is a question. He kept asking it. “Why should one be limited to the number of careers and opportunities just because they are living with a disability?”
It is the kind of question that, if you can answer it, gives you a company.
He came back to Nairobi in 2023 and started Zerobionic.
He started it with a co-founder — Norah Kimathi, a fellow Strathmore student, who would still be in her fourth year of undergrad in the most recent press. He started it inside a Strathmore robotics lab, with university-provided space and mentorship from the university’s two innovation arms, @iBizAfrica and @iLabAfrica. He started it as the CEO, the hardware engineer, and the machine-learning engineer, all three roles at once, because there was nobody else to be them.
And he started it, deliberately, as a disability-led company. Not a robotics company that happened to do disability work. A company founded by, and explicitly for, people with disabilities. In an investor culture that often nudges founders to abstract themselves away from their own story, that choice was unusual. It would also turn out to be a defence — and a thesis — and, eventually, a strategy.
The first robot was not very good. He called it Model 4. It was a hand. It moved. It signed. Not well. Not yet.
But it was a start. And the deeper insight — the one that would carry the company — was already in place.
A piece of hardware was being built, on a student budget, by the very kind of person it was for.
Eight Hands Between Two Models
Launching a humanoid robot is not like launching an app. There is no minimum viable product you can deploy to a few thousand users overnight. Every iteration is physical. Every iteration costs money the team did not have. And every iteration has to actually work, in a real classroom, in front of real children, on the first try, because there is no second chance with a deaf student who has been let down a hundred times before.
Zerobionic’s first hard problem was simply the hand.
A human hand has twenty-seven bones, dozens of articulated joints, and a vocabulary of micro-movements that human evolution spent millions of years tuning. Recreating that in plastic, at student-budget scale, with servos and tendons and 3D-printed knuckles — for a sign language that uses both hands and the spaces between them — is one of the harder things you can ask a small team to build.
Zerobionic was building it, deliberately, out of recycled plastic waste. More than sixty per cent of every robotic system the company makes is built from plastic that was, until recently, somebody else’s garbage. The company’s own framing is that its circular manufacturing process cuts industrial waste by up to ninety per cent. It is a beautiful number. It is also a constraint. Recycled plastic is harder to work with than virgin polymer. It is less consistent. It warps differently in the print bed. It snaps in places virgin plastic does not. And it forced the team to engineer around the material every step of the way.
The second hard problem was the language.
Kenyan Sign Language — KSL — is a rich, expressive, living language. It is also, like most African sign languages, under-resourced. It has limited published vocabulary for the kind of words a STEM class needs every day. Calculus. Organic chemistry. Electromagnetism. Cellular respiration. There are simply not enough signs.
So Zerobionic could not just buy a dataset and train a model. There was no dataset. The team had to co-develop the signs themselves, with deaf educators, on the ground, in workshops — invent the sign for “isomer,” agree on the sign for “covalent bond,” validate them with the deaf community, and only then teach the robot. The robot was learning a language that, in the technical register, did not yet fully exist.
The third hard problem was the customer.
Maxwell was twenty years old, walking into the offices of school principals and county education officers, asking them to put a robot in their classroom. Many of them, understandably, had questions. They had also never seen the technology. They had never met another school that had it. They could not pick up a phone and call a peer for a reference. Every pilot meant winning over a sceptical institution from a cold start.
And the fourth hard problem was the one Zerobionic chose for itself — and refused to give up. The disability-led identity.
In African deep-tech, foreign capital often prefers a founder it can package in a familiar way. A Stanford degree. A Y Combinator stamp. A clean, abstract pitch. Maxwell’s story — Nakuru kid, prosthetics outreach, disability-led mission — was not that. And he resisted being smoothed into it. The mission was the company. He would not edit out who he was to make the pitch deck easier.
So the team did what small, stubborn teams do. They built. Model 4, then 5, then 6. They tested each one in a real classroom, not a simulation. They treated deaf students not as test subjects but as design partners — the people who would tell them, kindly and unsparingly, that the index finger on the new hand was bending half a degree wrong, and that this changed the meaning of an entire sign.
And somewhere across that grind, two things happened that pulled the company out of the struggle and into the next chapter.
The first was that the robot got good enough to demonstrate publicly. The second was that the world started watching.
The Twenty-One-Year-Old Who Said No to VC
By the end of 2024, Zerobionic had stopped looking like a student project and started looking like a company the world was tracking.
In 2024, the company won the Gold Award for Access, Diversity and Inclusion at the QS Reimagine Education Awards — a global competition for educational innovation. In June of that year, Maxwell was on a stage in Switzerland, presenting at the Global Learning Council’s annual meeting, after winning the GLC’s hackathon hosted by Strathmore and HP. He was accepted into the Global Learning Council Fellowship. By November, Zerobionic was one of just ten African startups selected for the Latitude59 Kenya pitch competition — the Kenyan regional round of a pan-European tech competition with a one-million-euro prize pool at its Tallinn finals.
And then, in February of 2025, came the validation that mattered most at home. At the Africa Tech Summit in Nairobi, Zerobionic won the Young Tech Innovator Award. A speech-to-sign motorised prosthetic arm, built by a Strathmore undergraduate from recycled plastic, was named the most promising young innovation on a continent of more than a billion people.
And here is where the company’s defining strategic choice gets named out loud.
Sometime in mid-2025, Maxwell sat down on a Nairobi podcast called Mission Accomplished, produced by an organisation called Mtoto News. The episode is on YouTube. The headline they put on the thumbnail is the one that travelled. “I’m a Millionaire at 21.”
And we need to sit with that line, because this is Asili, and honesty is the brand.
What Maxwell said, and what the Kenyan press faithfully reported, is that he had become a millionaire by twenty-one. In Kenyan shillings. One million Kenyan shillings is — at mid-2026 exchange rates — somewhere around seven thousand seven hundred US dollars. It is a meaningful, hard-won personal milestone for a Nakuru kid who started with nothing. It is not the “USD millionaire” image that an English-language headline can imply. We are going to call it what it is. A KES millionaire. A Nakuru kid who broke through. Not a Forbes cover.
But on the same podcast, he said something more strategically interesting than the headline. He said he had built Zerobionic, deliberately, without venture capital.
No funding rounds. No Crunchbase entry. No equity sold to a foreign fund. Instead, the company had been built on prize money, university-provided infrastructure, mentorship grants, and whatever revenue the early pilots generated. The capital strategy was a thesis as much as a constraint. Maxwell wanted to keep the mission disability-led — to keep the company answerable to the people it served, not to a fund that needed a ten-times return inside seven years. That is a hard road. He chose it anyway.
By the end of 2025, Zerobionic had pushed the hardware from Model 4 to Model 12. The accuracy figure the company began quoting was ninety-two per cent translation accuracy — by its own measure, with no independent benchmark yet published. The classroom pilots were live in three counties — Garissa, Machakos and Kirinyaga — running through a partnership with an organisation called Young Scientists Kenya.
A robot built from trash, in a university lab, by a 21-year-old without venture capital, was now signing chemistry to deaf students in three Kenyan counties.
What happened next was the world deciding it could not look away.
One of Ten, Out of 1,200, From 45 Countries
January 2026. The Africa Tech Summit Nairobi names its annual Investment Showcase cohort — twelve startups, selected from the entire continent, set to pitch in front of investors. Zerobionic is one of them.
Three months later — the ninth of April, 2026 — the announcement that puts Zerobionic on a different scale entirely.
Qualcomm — the American chip giant — runs an annual mentorship programme called Make in Africa. It is one of the most competitive deep-tech programmes on the continent. In 2026, it received more than twelve hundred applications, from forty-five African countries. From that field, it selected ten startups. One of them, the only Kenyan company on the list, is Zerobionic.
The programme is equity-free. The companies in it receive technical mentorship from Qualcomm’s engineers, intellectual-property training delivered through a programme called L2Pro Africa, a completion stipend of five thousand US dollars, reimbursement of up to another five thousand for filing patents, and eligibility for grants from the Qualcomm for Good Social Impact Fund. For a hardware company that has refused to take dilutive capital, this is precisely the kind of validation that does not cost a single share.
And it is the validation that travels. Once Qualcomm names you the only Kenyan startup in its Africa cohort, the rest of the audio industry — and the rest of the African venture press — pays attention.
In May of 2026, just a month after the Qualcomm announcement, Zerobionic is named to the Africa Tech Summit London Investment Showcase — thirteen African startups travelling to London to pitch to international investors. The framing of that London showcase is explicit. The companies presenting are seeking partnerships to scale production and expand deployment. Zerobionic is no longer in stealth. It is no longer just iterating in a Strathmore lab. It is on a stage in London, looking at the global capital markets, with a product the global capital markets have not seen before.
Behind these announcements is a less visible second story. The institutional relationships are deepening. NVIDIA — the company that makes the chips that run most of the world’s AI — has been credited in African business press as providing global connections and support, the form of which has not been publicly described. Strathmore continues to anchor the lab. The @iBizAfrica and @iLabAfrica centres continue to anchor the mentorship.
A robotics company that began with a single 3D printer and a student’s stubbornness now has Qualcomm’s name on its mentor list, NVIDIA on its acknowledgements, Strathmore on its lease, and an investor day in London on its calendar.
But scale, for a hardware company, is the moment where the previous chapter’s defining choice — no venture capital — meets the next chapter’s defining demand: production. Manufacturing. Geographic expansion. Sign languages beyond KSL. Servos and chips and supply chains.
And that is where Zerobionic’s story sits today. Right on top of a question nobody has answered yet.
Will the No Become a Yes?
In the middle of 2026, Zerobionic is a company in the open part of its arc. Model 12 is the current generation of the robot. The Garissa, Machakos and Kirinyaga pilots are still live. The Qualcomm cohort is running. Maxwell is in his final year at Strathmore, finishing the degree and the company at the same time.
The product is improving. The signed STEM vocabulary is widening. The team is asking what comes after a humanoid robot — whether the next generation should still be a full humanoid hand, or whether a screen-and-camera “lite” version could put the same translation in more classrooms at lower cost. They have not answered that yet, in public.
The investor question is the loudest one. The London showcase was framed explicitly around seeking partnerships. The no-VC stance that defined the first three years is, by all reading, softening. It may not break. It may evolve into impact investors, strategic partners, government procurement, a Qualcomm Social Impact Fund grant — capital structures that hold the mission in place. But the categorical “no” is becoming a “from whom.”
The geographic question is the next one. KSL is not the same as Ugandan Sign Language. It is not the same as Tanzanian Sign Language. African sign languages vary by country and even by region. Scaling Zerobionic across East Africa means re-training the machine-learning models for each new language, in partnership with each new deaf community — a different and harder challenge than just shipping more units.
And the harder question, the philosophical one, is the one that closes this story. Africa has, by Zerobionic’s own framing, around thirty million deaf and hard-of-hearing people. For most of modern history, deep-technology built specifically for them — and by them — has come from somewhere else. Western universities. Western non-profits. Western design assumptions about what disability looks like. Zerobionic is one of the first deep-tech assistive-hardware companies on the continent that was founded by, and built for, the people it serves — and that is choosing, deliberately, to keep it that way.
The empire it is building is small, still. A handful of classrooms. A handful of counties. A robot the size of an arm. But the gesture is enormous. A Nakuru kid asked why disability should set a ceiling on a life — and built a pair of hands to lift it.
Every empire has an origin.
Zerobionic’s origin is a prosthetics clinic in West Africa, where a 19-year-old from Nakuru noticed who was being left out of the conversation — and decided to build the hands that would let them back in.
It has not become the loudest company in African tech. It is becoming one of the most necessary. And the question it leaves us with is the only one that ever really matters in this work: who gets to build the future, and who gets to be in the room when it arrives?
This is Asili Africa.
Key Takeaways
- The Nakuru Kid Who Went to Ghana. To understand Zerobionic, you have to start in Nakuru.
- The Twenty-One-Year-Old Who Said No to VC. By the end of 2024, Zerobionic had stopped looking like a student project and started looking like a company the world was tracking.
- One of Ten, Out of 1,200, From 45 Countries. January 2026.
- Will the No Become a Yes?. In the middle of 2026, Zerobionic is a company in the open part of its arc.
Also available on YouTube — search “Asili Africa” or subscribe to our channel.