Investor FAQ

Our USP is not a single feature but a combination that no competitor currently has, and which is hard to copy because of how the company is built.

Completeness. A blind person’s day is not one task. They need to cross a road, avoid a low-hanging branch, read a letter, find a door, recognise a friend and, when something goes wrong, reach a human. Existing products cover one or two of these: Envision and OrCam read text and recognise objects; Meta and Apple sell mass-market glasses with no obstacle detection and no support for a blind user; Be My Eyes and Aira offer a remote helper but no hardware. Mavis covers all of them in one pair of glasses, with built-in SIM and GPS, so the user does not need a phone in one hand, a cane in the other and an app subscription on top. The obstacle-detection layer — depth and light sensors combined with the camera, protected by UK patent GB2410710.4 — is something the text-reading products do not attempt.

Price. At £1,230 Mavis is two to four times cheaper than OrCam and Envision. That is not a loss-leader; it comes from owning the whole stack. We design the PCB, firmware, software and frames in-house and assemble in Northampton, so there is no white-label hardware margin and no distributor margin. Hardware gross margin is 84% before channel commission, and the £30 a month subscription funds the Support Centre and ongoing AI updates.

In-house design and manufacturing. We are the only UK company that both designs and manufactures smart glasses for blind and visually impaired people, with CE marking and EMC testing complete. When a user reports a problem, the engineers who can fix it are in the same building, and improvements ship over the air. Competitors that depend on external hardware partners iterate in product generations; we iterate in weeks.

Human backup. The Support Centre is a trained, DBS-checked team reachable from a button on the glasses, not a volunteer app. Users tell us this is the feature that makes them trust the technology enough to leave the house with it, and it is a recurring-revenue line that competitors selling a one-off device do not have.

Built by a user. Our founder has lived with severe visual impairment since childhood and is the company’s CTO. Product decisions come from lived experience, and the sight-loss community — RNIB, Guide Dogs, Sight Scotland, the Northamptonshire Association for the Blind — has engaged with us as one of its own. That credibility is very hard for a large consumer-electronics company to buy.

Privacy. The glasses cannot record video or store images, all traffic is encrypted, and all processing stays on UK servers. For a user who is, in effect, wearing a camera all day, and for the NHS, charity and DSA channels who buy on their behalf, this matters.

Put together: the most complete product, at the lowest price, from the only UK company that owns its hardware, backed by humans, designed by someone who needs it. Any one of those could be copied; the combination is the moat.

When we started, the obvious route was to write software for someone else’s glasses. We tested that route and it fails on three counts.

The sensors are not there. Detecting a low branch, a step down or a bollard needs depth and light sensing fused with the camera, and it needs to run continuously and locally, not through a cloud round-trip. No consumer frame has those sensors, and no maker of consumer frames will add them for a market of a few hundred thousand users. Our patent (GB2410710.4) is on exactly this triple-sensor arrangement.

The interface is wrong. Mass-market glasses assume the user has a phone in their hand and can see a screen. A blind user needs tactile buttons, audio that leaves the ears free to hear traffic, haptic alerts, and a device that works with no phone at all. Mavis has built-in SIM and GPS for that reason.

The economics are wrong. Building on someone else’s hardware means paying their margin, accepting their release schedule and living with the risk that they change the API or discontinue the product. Owning the design gives us a unit cost of around £200, an 84% hardware gross margin before commission, and the freedom to change anything within weeks.

There is a fourth reason that matters to investors: the hardware is where the defensible IP sits. Software features can be copied in months. A patented sensor arrangement, a certified electronic design and a working production line are much harder to reproduce.

Every investor asks this, and it is the right question. The answer has three parts.

Different customer, different product. Ray-Ban Meta is a camera, speakers and an AI assistant in a stylish frame, controlled from a phone. It will describe a scene if asked. It will not warn you about the kerb in front of you, because it has no depth sensor, no sensor-fusion running continuously, and no reason to add either for its customers. Apple, Google and Samsung are on the same path: displays, cameras and assistants for people who can see. The features a blind person depends on most are the ones a sighted consumer never uses.

Big Tech has told us how they think about this market. Meta’s own accessibility features are delivered through partnerships with Be My Eyes and Aira, volunteer and paid remote-assistance services running on top of the glasses. That is the model: the platform stays general, and accessibility is handed to third parties. Mavis is one of the companies that specialised product will come from, and the more Big Tech normalises glasses on people’s faces, the easier our sale becomes.

Where we win is exactly where they will not go. Human support staff with DBS checks, UK data residency, no recording, DSA and NHS procurement compliance, CE marking as an assistive device, a support line staffed by people who understand sight loss, and a founder who is a user. None of that scales to two billion customers, so none of it will be built in Menlo Park or Cupertino.

Two honest caveats. First, if Big Tech ever adds real obstacle detection to a consumer frame, that would confirm the market rather than end it; the assistive layer, the support and the procurement relationships would still be ours to sell, and we would be an obvious acquisition. Second, we do compete with them on price at the low end: a £379 Ray-Ban Meta is enough for some people with mild sight loss. Our market is the 277,000 people in the UK with severe sight loss and their equivalents worldwide, for whom it is not.

The founder is the user. David Sikharulidze has lived with severe visual impairment since childhood and is our CTO. Every product decision is tested on the person making it. The sight-loss community can tell the difference between products designed for them and products designed with them, and it has responded to Mavis accordingly.

Community trust. RNIB, Guide Dogs, Sight Scotland and the Northamptonshire Association for the Blind engage with Mavis as a partner. RNIB’s Robin Spinks appeared alongside us on Sky News. RNIB Tech Talk ran an episode on user experiences with the glasses. That endorsement is not for sale.

Certified for this use. Mavis is CE marked and EMC tested as a device for blind and visually impaired people, which is what NHS trusts, DSA assessors and charities need to see before they can recommend or fund it. Consumer glasses are certified as consumer electronics; that is a different thing.

The procurement channels. Being an approved supplier through Microlink under Disabled Students’ Allowance, and eligible under Access to Work, means the state pays for Mavis on behalf of the user. Those channels are built by assessors, paperwork and relationships over years. A Ray-Ban is not an assistive device and does not go through them.

Human support as part of the product. The Support Centre is a trained, DBS-checked team, on a button on the glasses, funded by the subscription. A platform company will always route this to a volunteer app or a partner. We employ the people.

Speed and focus. We design, build and support the product in one building in Northampton. A user reports a problem on Monday; the fix ships over the air the same month. For a large company this market is a rounding error; for us it is the whole company.

Selling to blind people has a specific problem: the buyer often cannot afford £1,230, but there are well-funded schemes whose purpose is to buy this kind of equipment for them. Our go-to-market is built around those.

1. Disabled Students’ Allowance (live). Mavis is available in England through Microlink, a DfE-appointed DSA assistive-technology supplier. A student is assessed, the assessor recommends Mavis, and the government pays. The equipment cap is well above our price. We are extending this to Scotland (SAAS) and Northern Ireland (SFNI) in 2027, and to Irish colleges through the Fund for Students with Disabilities.

2. Access to Work. The DWP scheme pays for equipment that lets a disabled person do their job. In 2023/24 it funded 67,720 people and spent £257.8 million, and equipment was the fastest-growing element. Mavis fits squarely within it; we support applicants through the paperwork.

3. NHS and charity partnerships. Low-vision clinics, eye-care liaison officers and local sight-loss charities are the trusted point of contact for someone newly diagnosed. Pilots are planned for 2027, and the charities we work with already refer users to us.

4. Direct and Ambassador sales. The website, accessibility exhibitions (Sight Village, NAB, Seescape) and media coverage produce direct enquiries. The Ambassador programme pays commission to individuals, content creators and organisations who introduce buyers; we handle the sale, delivery, support and warranty. The financial model assumes 20–25% of revenue goes to channel commission, and still leaves a contribution of £631–£692 per unit.

From 2027 a dedicated sales team takes over the channel work from the founder, and from 2028 the same playbook runs in Ireland and then the EU, where the CE mark already allows sale.

Point by point:

Obstacle detection. Mavis: yes, continuous, on-device, adjustable one to three metres, patented triple-sensor design. Meta: none. This is the single feature blind users rank highest, and it is the one Meta cannot add with a software update.

Works without a phone. Mavis has built-in SIM and GPS. Meta requires a paired smartphone and an app, which many older blind users do not have or cannot operate.

Controls. Mavis has tactile buttons and haptic feedback. Meta relies on a touch strip and voice.

Human help. Mavis: a trained, DBS-checked Support Centre on one button, included in the subscription. Meta: third-party apps (Be My Eyes volunteers; Aira at additional cost).

Privacy. Mavis cannot record or store images and processes everything on UK servers. Meta glasses record video and photos by design, which is a problem in workplaces, hospitals and schools, precisely where our users need help.

Funding. Mavis is an assistive device eligible for DSA and Access to Work. A Ray-Ban is not.

Where Meta is better: fashion, brand, battery life for music and video, and price. For someone with mild sight loss who wants scene description on holiday, it may be enough. For someone who needs to cross a road alone, it is not the same product.

United Kingdom. RNIB estimates more than two million people living with sight loss, projected to double to over four million by 2050. Around 340,000 are registered blind or partially sighted, split roughly half and half. Of the two million, 277,000 (13%) have severe sight loss, 469,000 moderate and 1.35 million mild. About 26,000 are children. Our core UK market is the severe group plus the more affected part of the moderate group: roughly 360,000 people, of whom the 2031 plan reaches 14,000 cumulative units, under 4%.

Europe. The European Blind Union estimates 30 million blind and partially sighted people in geographical Europe, about one in thirty, with four partially sighted people for every blind person. Over 75% of those of working age are unemployed, which is why state-funded assistive-technology schemes exist across the EU. The CE mark means Mavis can already be sold in every member state; the plan enters Ireland in 2028 and the wider EU through distributors from 2029.

Worldwide. The IAPB Vision Atlas (Lancet Global Health Commission, 2020 data) counts 43 million blind people and 295 million with moderate-to-severe vision impairment, out of 1.1 billion with any vision loss. By 2050 those become 61 million and 474 million. Prevalence is highest in South Asia (9 million blind in India alone), North Africa and the Middle East. The assistive-technology market as a whole was £14.2 billion in 2023, growing at 8.5% a year to £23.7 billion by 2030.

The point for investors: the UK alone supports the whole six-year plan. Europe and North America are upside, not assumptions.

DSA is a non-repayable grant from Student Finance that pays for equipment and support a disabled student needs to study. For a blind student, an assessor recommends the equipment; a DfE-appointed supplier (Microlink or Study Tech in England) supplies it; the government pays the supplier. The student never sees an invoice.

The numbers: the House of Commons Library reports around 72,000 DSA recipients a year in England at a cost of about £107 million, and HESA data shows roughly 3,190 students declaring blindness or a serious visual impairment in 2019/20. Because a degree lasts three to four years, that is around 800 to 1,000 new blind or seriously visually impaired students entering higher education each year in England, plus a proportion of the far larger low-vision group who declare a different primary disability.

What matters commercially is that Mavis is already inside the system. We are an approved product through Microlink, and £1,230 sits comfortably under the equipment allowances (Scotland’s cap alone is £5,160; Northern Ireland funds up to £25,000 a year). A DSA sale has no price objection, no credit risk and a customer who will use the product every day for years and then carry it into work, where Access to Work takes over.

The 2031 plan needs 14,000 cumulative units across all channels. Capturing one in four new blind or seriously visually impaired students in England alone would contribute over 1,000 units across the plan, before Scotland, Wales, Northern Ireland and Ireland.

Access to Work is a DWP grant that pays for the extra costs of a disability at work: equipment, support workers, travel and adaptations. It is available to employees, the self-employed and people about to start a job, and it is not means-tested. Equipment such as Mavis is funded outright.

The scheme is large and growing: 67,720 people had provision approved in 2023/24, up sharply on the previous year, with total spend of £257.8 million. The “special aids and equipment” element was approved for 41% of recipients and more than doubled year on year. The DWP does not publish a breakdown by sight loss, so we estimate the addressable group from RNIB data: roughly 85,000 registered blind and partially sighted people of working age, of whom only about a quarter are in work. RNIB notes that the employment rate for blind people has not improved since 1991.

That gives two markets. The first is the roughly 20,000 blind and partially sighted people already employed, for whom Mavis makes the commute and the workplace navigable and who can apply today. The second, larger one is the 60,000 or so not in work, for whom Mavis plus Access to Work is part of the route into employment; charities and Jobcentre disability employment advisers are the referral path.

Our part of the process is small but important: we supply the quote and specification an applicant needs, and our Support Centre is a line item the scheme can fund alongside the device.

Country figures are approximate and drawn from national blindness organisations and census data; registration systems undercount because many people never register. The regional totals are those used in the financial model. We are happy to share the underlying sources on request.

CE marking is a legal declaration that a product meets every applicable EU directive, and the same standards apply for UKCA in Great Britain. For Mavis the applicable set is broad: the Radio Equipment Directive (the glasses contain 4G, Wi-Fi and Bluetooth radios), electromagnetic compatibility, electrical safety including the lithium battery and its charging, RoHS on materials, and specific-absorption-rate limits because the device sits against the head.

Each of those requires accredited-laboratory testing of the final production design. Change the antenna, the PCB layout or the battery and parts of it must be repeated. A small company building on a pre-certified module can inherit some of the module’s radio approval, but only if it does not touch the antenna or the enclosure, which a glasses design almost always does. Companies that assemble prototypes from development boards, or that ship early units to users as “trials”, are often doing so precisely because they have not got through this. Some sell into the UK relying on the user to import; that route is closed to any institutional buyer.

Why it matters commercially: DSA suppliers, NHS trusts, charities and retailers will not list a product without the mark, and neither will Amazon. Distributors and importers carry legal liability for non-compliant goods, so they refuse them. Insurers ask for it. And the CE mark is the passport to selling in all 27 EU member states without further approval; our EU authorised representative is already in place in Vienna.

Mavis passed EMC and CE testing on the production design in 2026. Because we own the design and the manufacturing, we can plan hardware revisions around re-testing rather than being blocked by it. Whether Mavis is classified as a medical device is a separate question we answer below.

Margin. Our unit cost of about £200 is built from 146 priced line items. An outsourced assembler quoted £99 per unit for the PCB alone at 200 units; at our volumes, contract manufacturing would add cost, not remove it, and would take a share of every unit forever.

No minimum orders, no inventory risk. Contract manufacturers want runs of thousands. We build in batches matched to demand, so cash is not tied up in finished stock and the plan’s working capital is small.

Speed of iteration. A design change goes from the engineer’s desk to the assembly bench in the same building. We have shipped several hardware refinements since May based on Pathfinder feedback; a factory in Shenzhen would have turned each into a three-month project.

Quality and calibration. Every pair is tested and its sensors calibrated by us before it ships. For a safety-relevant device used by people who cannot check it visually, that matters, and it is what the CE technical file requires us to control.

Service. Repairs, battery replacement and hardware upgrades happen here, with 14-day turnaround. Users are not waiting on an overseas RMA.

IP stays inside. Nobody outside Mavis has the full design files, firmware and test procedures. That protects the patent position and makes the company a cleaner acquisition.

Made in the UK. It is a procurement advantage with public bodies, it supported our Innovate UK and IPO recognition, and it is a genuine reason customers and press have engaged with us. The Manufacturer and Manufacturing Management both covered the launch on that basis.

The facility is already equipped and operational; this round funds people and stock, not machinery. Capacity today is several hundred units a month on one shift, enough for the plan to 2028, and the constraint is component lead time rather than assembly.

The Medical Devices Regulation applies to products with a medical purpose: diagnosis, prevention, monitoring, treatment or alleviation of disease or injury. Mavis compensates for a disability in daily life; it does not claim to treat the eye condition. That places it alongside white canes, screen readers and mobility aids rather than alongside implants and diagnostic imaging, and regulators in the UK and EU treat assistive technology of this kind as general consumer/electrical equipment.

We are careful in how we describe the product for exactly this reason. Mavis helps a user detect obstacles and navigate; it is not marketed as a replacement for a cane, a guide dog or clinical care, and the user manual says so. Our advisory board includes regulatory experience, and we review claims before they are published.

If a future feature crossed into medical territory, for example measuring a clinical parameter, we would assess MDR classification for that feature specifically. Nothing in the current roadmap requires it.

The technology arrived. Vision-language models that can read a label, describe a room and recognise a face now run fast enough to be useful through a pair of glasses, and the inference cost has fallen to a few pounds a month per user. Depth and light sensors that were laboratory parts five years ago are now small and cheap enough to sit in a frame. Our subscription economics (£5.49 to deliver £25 of net revenue) only work because of this.

The category was validated. Big Tech has spent more than $10 billion on smart eyewear. The public now expects to see people wearing camera glasses, and blind users no longer stand out for it. That removes the biggest social barrier assistive wearables used to face.

The funding channels are in place. DSA, Access to Work and the equivalent EU schemes will pay for assistive technology today. We are already inside one of them, with the others in progress.

The incumbents are weak. OXSIGHT, the previous UK player, went into liquidation. OrCam and Envision are expensive single-function devices. There is no established leader in complete assistive glasses, and there will be one within a few years. This round is what lets Mavis be that company rather than watch someone else become it.

Production. Risk: demand outruns a small assembly operation, or a key component goes on allocation. Mitigation: the equipment is owned, capacity is several hundred units a month on one shift, orders ship in 14 days, and we hold safety stock of long-lead components. Second-sourcing of the main modules is in progress.

Adoption pace. Risk: institutional channels (DSA, NHS) move slowly and direct sales alone cannot hit the plan. Mitigation: four channels running in parallel, the DSA channel already live, and a plan that reaches EBITDA-positive at 1,800 units, a small fraction of the UK market. The 2026 conservative case (56 units) is modelled and funded.

Competition and IP. Risk: a large player copies the approach, or a competitor challenges our patent. Mitigation: UK patent granted, PCT application published, six registered designs, and the sections above on what cannot be copied. We chose to own the hardware partly so the IP is real rather than a software layer.

Capital. Risk: the round takes longer than planned or the cash trough is deeper. Mitigation: the plan needs £399,000 to reach profitability and this round targets £500,000, a £100,000 buffer; the company runs lean, has secured £160,000 or more in grants to date, and qualifies for EIS. The model also shows a route to profitability with no further external capital after this round.

Key person. Risk: the founder is also the CTO and the face of the company. Mitigation: hiring in 2026–27 is aimed at exactly this; two support staff join in October, the sales lead in 2027, and the advisory board provides continuity. Founder key-person insurance is planned as part of this round.

The assistive-technology sector consolidates regularly, and buyers pay for three things Mavis is building: a certified product with IP, a recurring-revenue user base that is expensive to acquire and rarely churns, and access to institutional channels. Likely acquirers fall into three groups: assistive-technology groups looking to add a wearable to a portfolio; med-tech and eye-care companies that already sell into the same clinics; and eyewear or consumer-electronics companies that want a credible accessibility line without building the trust from scratch.

The timing follows the plan. Break-even in 2028 and cash-flow positive in 2029 remove the need to sell under pressure; European roll-out in 2028–29 creates the multi-country footprint acquirers want; and by 2031 the model shows £22.4 million revenue and £6.5 million EBITDA. At 9–15 times EBITDA, which is the range comparable transactions have paid, that is £59–98 million.

A trade sale is the base case, not the only case. The business is designed to be profitable on its own, so holding for dividends or a later, larger sale are both open. What we will not do is depend on a sale to survive.