Fuad Murtuzayev is Principal at Mount Fund and Founder of Keane Tech, a B2B computer vision platform for gym operators. He spoke with us about why fitness economics broke for hardware, what makes Keane’s software model structurally different, and why London is the right launchpad.
You wear two hats: Principal at Mount Fund and Founder of Keane Tech. From an investor’s seat, what made the fitness category compelling enough to cross the line from backing to building?
For me, the move from backing companies to building one was a question of when, not if. I recently reached the stage where I had both the freedom and the genuine desire to build something from scratch and be hands-on in doing it. Fitness was simply the category where that instinct and the opportunity finally lined up. At some point the analysis has to give way to execution.
The unit economics of consumer fitness hardware have historically been brutal — thin margins, high churn, commoditisation pressure. How does Keane’s model differ structurally?
This is exactly the trap we designed the model to avoid. Those harsh economics are largely a function of selling physical devices into a fickle consumer market. Keane doesn’t sell hardware, at least for now — we’re a B2B software layer, so there’s no device to manufacture, subsidise, or replace. That means three things.
First, our margins are software margins, so deployment doesn’t depend on shipping a product. Second, our churn profile is institutional rather than consumer: we’re retained by gym operators on long-term recurring licensing agreements, not by individuals who cancel after a few weeks. And third, because the value we deliver accrues to the operator’s bottom line, the product becomes stickier the longer it runs rather than commoditising.
Walk us through the founding insight. What convinced you the gym floor was specifically broken?
The gym floor has seen no shortage of innovation around consumer habits and training methodology, but almost all of it targets the workout itself — the exercises or the equipment. What we chose to focus on is fundamentally different: the process that envelops a person’s entire fitness regimen. Tracked sessions, automatic activity logging, and form rating all create an incentive to show up — they let your mind rest so you can focus on your body. That was the gap we saw, and it was structural rather than cosmetic.
Investors increasingly talk about “physical AI” as the next frontier after generative AI. Where does Keane sit within that thesis?
We don’t anchor ourselves too tightly to whichever label happens to be fashionable; the test for us is practical, not semantic. Keane provides a service to gyms, and ultimately to their clients, who save time and money while we smooth out the logistics of training for them. If “physical AI” means applying machine intelligence to the physical world in a way that produces those concrete outcomes, then that is squarely where we sit.
What gives the technology genuine defensibility beyond simply being early?
Being early is an advantage, but it isn’t the foundation of our defensibility — the compounding is. Every customer we sign gives us more real-world data to train on, so our computer vision improves with each deployment, much like compound interest. Underneath that sit three engineering strengths.
Modularity of our CV system. Our pipeline runs up to seven computer vision models that we improve simultaneously. Because they’re decoupled, improving one part of the system doesn’t block or destabilise the others during training.
In-house labelling and training. We’ve established our own labelling team that works directly with incoming client data to produce fresh labelled datasets for new challenges in computer vision — we don’t depend on third parties to teach our models.
A genuine data feedback loop. Customers give us data, we train, the model gets better, and the better product earns more data in return. Because we stay in close, accurate communication with our clients, we can raise exercise-recognition accuracy from real-life data in an extremely short window.
Together, these mean our edge widens with scale rather than eroding.
From an operator’s perspective, what’s the financial case for installing Keane?
Keane drives ROI across three vectors: improved member communication through automated session logging and progress updates that keep members engaged; stronger connectivity between gym staff and clients through real-time performance data; and higher conversion of casual gym-goers into committed, paying long-term members. Gyms that can demonstrate measurable progress to members retain them longer — and retention is where gym economics are won or lost.
Retention, pricing power, ancillary revenue — where does the ROI actually come from?
It’s simply smart business. People who use our service will almost inevitably want and need the data that overlaps into the wellness aspect — recovery, overtraining signals, the broader health picture. These things go hand-in-hand, so we offer clients an all-in-one product that covers most, if not all, of what they truly need, rather than forcing them to stitch together separate tools.
You’ve built recovery and overtraining analysis into the platform. From a business angle, why extend into the health and longevity layer rather than staying focused purely on tracking?
We built for the full spectrum of users — not just beginners, but serious athletes and those at genuine risk of overtraining. Recovery and overtraining signals aren’t a distraction from our core; they’re essential to it. Longevity is where the fitness industry is heading, and operators who can offer that layer will command premium positioning. More data also means a better product — every additional health signal we capture makes our recommendations smarter and our platform stickier.
London is the launchpad. How does the European market shape your commercial and capital strategy compared to scaling out of the US?
London gives us a dense, sophisticated market to prove the model with institutional gym operators before the capital-intensive push into the US. Europe’s regulatory environment also forces privacy-by-design discipline early, which becomes a competitive asset globally. We’ll pursue US expansion once unit economics are validated and we have referenceable partners.
Data privacy is the obvious risk question with any camera-based system. How do you address it from a governance and investor-confidence standpoint?
Privacy was built in from the start. We don’t keep the camera feed — the raw video only exists while it’s being processed, then it’s gone. All we store is the structured output: exercises, reps, counts, form quality. No images, ever.
That handles most of the risk on its own. No stored footage means nothing to leak or get subpoenaed, so the privacy problem that trips up most camera-based systems never builds up in the first place. We hold data about the workout, not pictures of the person.
You’re preparing pilot launches with gym partners. What does the next 12–18 months look like commercially?
We are preparing pilot launches with paying gym operators live on the platform within the next few months, then scaling across the UK to build referenceable case studies and prove the revenue model at volume. This sets up a clear path into broader European expansion and a Series A raise.
We have already been closely partnered with Olympia Personal Training and So Gym throughout development, meaning we go into this phase with established relationships and real operator insight already baked into the product. This phase will also allow us to fine-tune the platform with real-world usage and validate our research in a live commercial environment on a larger scale.
Final question — what’s the broader lesson from building Keane?
The most durable opportunities often aren’t about inventing something entirely new, but about re-architecting an existing problem so its economics finally work in your favour. Fitness had been approached as a hardware problem for years, and the economics punished nearly everyone who tried. We didn’t set out to build a better device — we set out to remove the device from the equation and deliver the same value as software, on infrastructure that was already there. For investors and operators, the takeaway is to interrogate the structure of a market, not just its surface trends — because that’s usually where the real edge lives.
About Fuad Murtuzayev
Fuad Murtuzayev is Founder of Keane Tech and Principal at Mount Fund. Keane Tech is currently onboarding pilot gym partners across London.
Sponsored Feature
