Karthik Mahadevan is the co-founder and CEO of Envision, a company that helps the visually impaired to live more independently. He is passionate about designing tools that empower people and is an alumnus of Delft University of Technology where he completed his Master’s in Industrial and Product Design.
Q. Can you tell us about your journey and how you founded Envision?
I first came to the Netherlands to study Industrial design in Delft University, before which I was studying engineering in India. But I’ve always been fascinated by Design, the reason being that Design is all about solving problems. When I was studying in the design school (At Delft), I was always interested in picking up early-stage projects that start with the problem itself instead of picking projects with a technology or a solution already in place and the designer trying to figure out how to push it out.
Towards the end of my masters in Delft, I was trying to look for a problem statement to work on for my thesis. At that time, I had gone back to India for my winter holidays and was invited by a school for the blind in Chennai, to give a talk about what career opportunities which students like them could have going forward. I was there to talk to them as a designer and was explaining to the kids about what it means to design and how to approach it. I remembered explaining to the kids that a designer is just somebody who ‘solves a problem’. So if you’re able to identify a problem and if you’re able to put your skills and energy into designing a solution for it, you are also a designer!
Towards the end of the talk I asked a question to the kids in the audience that day - ‘If all of you could be designers tomorrow, what would be the kind of of problems that you would like to solve?’ Almost all the kids in the room that day said things like - “I want to be more independent”, “I want to be able to go and hang out with friends by myself”, “I want to be able to go to school on my own”, “I want to be able to pick up and read a book by myself”.
This desire for independence was such a strong emotion that they all want to experience, and it’s something that stayed with me. I thought - Why am I not doing something about this? I’m supposed to be a designer- somebody who solves a problem. Why don’t I explore this and see if there is a solution to this problem.
I came back to university and spoke to a professor about it, saying I wanted to explore this area as a problem statement for my thesis. I wanted to understand - what is independence? What exactly do they mean when they say they want to be more independent, what does independence mean to them? As I started to interact with similarly challenged folks here, I started to understand that for a lot of them, independence almost always meant ‘access to information’. This is because there is so much of information around us, and all of it in a visual format which leads to their inability to access it.
For example, when a blind person walks into a train station, the information is set up on a display. But because it is visual information, they cannot access it and they need to ask someone for help. In a way, it’s a shortcoming of society that most of our infrastructure has been designed in such a way as to be only visually accessible due to which, we are excluding a big segment of the population. At the same time though, I also understand that it’s pretty impractical to expect all the infrastructure to change overnight. Thus, I tried to approach this problem from another angle.
OK, all the infrastructure and information is going to stay the same, is there a way that technology can understand this information and have it be translated in an accessible way for the blind and visually impaired people?
I started exploring technologies like artificial intelligence, especially what they can do in terms of processing and extracting information from images and have them spoken out. A friend and I got together at nights to hack and explore the possibilities that we had with the AI technology of the time.
One of the very first things that we built was a very simple object recognition tool where it recognised simple objects. When we tested it with the blind and visually impaired people, they said, “OK, this technology is pretty cool but it’s of no use to us. If you tell me, there is a plate on my table, I’m probably the person who put the plate on the table. So it’s not useful information to me. It’s cool that technology can do it, but there isn’t a utility for me.”
That’s when things started to get interesting and we started to probe in new directions with respect to what information users are trying to look for and mapping it to a technology that can offer them the information. It’s not very interesting to recognise ‘a cup’ but it is interesting if I can recognise it as ‘your cup’.
Building such things is what added utility to our user instead of just building a journal with a technology aspect to it. From that point on, we spent our nights hacking a prototype on the basis of the feedback we received and then we’d go back the next day with with an updated prototype, for more feedback. We did this process a couple of times and towards the end of my thesis, we had built an MVP of the Envision app.
It allows a blind or visually impaired person to take images of various things and depending on what information they’re looking for, it extracts it and communicates it including capabilities like reading your text, finding objects, describing a scene, looking for particular faces (people) and a bunch of other things. We were able to package all of the technology in a way where a blind person can operate it on a smartphone without having to look at the screen.
So, that’s the point where it actually becomes effective because everybody appreciates just something that you have done. So when you build something, everyone appreciates it, and says you have done a good job but the real test of finding out value is when people are ready to pay for what you built. That’s when we understood that we have something amazing on our hands here.
If we were to put this in the hands of all the blind and visually impaired folks around the world as it was intended to be, we would have to build a business model around it.
We had to make sure it was sustainable and scalable. And that’s how we started to think about building a business around the idea. After we got accepted into a startup incubator here in the Netherlands, we launched our very first app on the App store and the Play Store in 2018. Since its launch, we now have tens of thousands of subscribers around the world who pay a subscription in order to use the app.
I think that’s the power of software, where you can put something out and though it’s not completely finished, it still adds value to some people. But the app was getting a lot of traction. It was very interesting but then one feedback that kept coming back to us, from our end users, was that the app was really cool, but it would be even better if they didn’t have to hold a phone in their hand, especially when they’re out and about.
We didn’t have a lot of experience in building hardware and we were clear from early on that we didn’t want to build our own. We started scouting for smart glasses and wearable cameras that are out in the market. We hit a roadblock with the available glasses in the market either from a technology standpoint or from an end user standpoint. A few of the glasses didn’t have the processing power to run our operations on them , the ones that were capable, were too big and bulky for a lot of our end users to actually have on the face.
For a lot of blind and visually impaired people, they were worried about drawing more attention to themselves. “I already attract a lot more attention when I’m out and about and if you put something that’s big and bulky on my face, I’m gonna attract even more attention and I don’t want that stigma attached to me.”
Towards the end of 2019, we won the Google Play award for the best accessibility app and they invited us to their office in California. They introduced us to the team that was working on the second edition of Google Glasses. Google Glass, launched a while back, but it didn’t do very well. It was too early for its time and people didn’t feel comfortable having it on their faces. So, internally Google started offering Google Glasses only for Enterprise applications.
They were about to launch a second edition of the Google Glass, which has better cameras, and processors. So we struck a deal with them where they offered us early access to this device. When we started prototyping with it, we found that it had a lot of the capabilities in terms of operating power that we needed. So it had a processor that was strong enough and at the same time, the form factor of these glasses were not as stigmatising as some other ones that we had been working with.
When we started to hand this out to our end users, they were okay with having this on their faces. It’s not unnoticeable, but it is something that is not a big, robotic device that you have on your face. We stuck a partnership with Google where we make a purchase of these glasses and entirely flash it with our software and this is something that we are now offering to end users as a product of Envision calling “Vision glasses”.
We were able to secure a hundred people who placed orders for them, which in turn increased the validation for the fact that there is a demand for something like this. To date, we have shipped 500 of these glasses to end users. And as of now, we’re at a phase where we are building up a distribution where stuff like these glasses can be offered to end users.
The stage at which Envision is in right now is - we have an app which already has a strong user base. We are constantly adding more value so that we can expand our user base and the glasses are a new addition for us, operationally. Hardware is not as easy to scale it as software. But the margins are certainly higher. So we can actually invest a lot more of the money that we make from the glasses to expand our businesses.
Right now, the idea is that all the revenue we make from the sale of the glasses, we’re going to invest into building a low cost of version of the same so that people in countries where there isn’t nationalised healthcare or there aren’t insurance available at a national scale, can also afford them
From a company standpoint, as of now, the focus for us is the blind and visually impaired, folks. But then in two years we want to expand this to also make applications for people with dyslexia or people with dementia or other cognitive disabilities.
We believe that in five years time, stuff like smart glasses, in-built cameras, powerful processors are going to become ubiquitous. Once everybody has access to such a camera on them and we are able to build intelligent software on it that can process images, which in turn allows us to offer more meaningful information that could better augment more people. So you want to go from building a vision based app for blind and visually impaired people to eventually enhancing vision for everyone and that’s a goal of the company.
Q: Speaking about accessibility, a lot of people speak about the solutioning side of its designers, what are some of the things that we can account for in the stage of user research and working with people who are different?
What we’re doing with Envision is a bit different because it is a product as of now only for the blind and visually impaired. But if you are attempting to build any sort of a product instead of attempting to make it ‘accessible’, I would say attempt to make it ‘inclusive’. And there is a subtle difference between accessible design and inclusive design.
Inclusivity means that you build ‘one thing’ that is for everyone. To give you an example here in Netherlands, everyone often takes public transport. They have an option that when a person with a wheelchair is going on the strains, you can, with an app, make a booking of which station you are going to exit and there will be a person at this station with a ramp that they put on the door of so that the person with a wheelchair, can exit as soon as they arrive at the station, now this is accessible design.
It is accessible for a person with a wheelchair, he/she is able to use the train, able to make it off the train, but it’s not that good because now the person still has to do something that everybody else does not have to do. This as opposed to what they’re doing now, where they’re ordering a bunch of new trains but the height at which the doors open has been modified. So they have a standard height for platforms and they made it a rule that all the trains that are ordered will have a door at a certain height, which means that there is no longer a need for a ramp. They are going to select doors which are going to open at the same level as your platform. This means that now a person with a wheelchair doesn’t have to ask for help. It’s inclusive now, because it is only one design that is for everyone.
So I think that’s the approach that you need to have, where anything that you are designing should be done in such a way that it is able to cover all extremes. It should never be an afterthought. So if you want to start inclusive design, it needs to be at the very beginning of the process itself.
And it’s not a lot of effort, you don’t have to change your design very much to make things inclusive. We are biased towards thinking that we need to make it officially less appealing (when designing for accessibility) because whenever we like to think about technology for old people, we always tend to think about phones with big buttons and chunky designs. I think it’s a major misconception that if you’re trying to make something inclusive it needs to start looking ugly or it needs to start looking big.
It is important that we are approaching it very openly, without having any inhibitions it. The right approach is to be more open and to be more upfront about it and to understand that it is possible to make something accessible and look good at the same time.
Q. How do we convince people to actually try a design solution and communicate it’s value?
I think the right way to approach it is it to get to the bottom of why do they feel that way? (Not wanting to give your design solution a chance). There’s generally a lot of assumptions for the user, when you’re designing something. When you say that building an app, they have the highlight and they already have a set of preconceived notions about what an app is. And I think that is something that is good to avoid early on when you’re still at the prototyping stage.
You just say that - I’m building a service! That’s probably a better term to use. It is a service that is offered to you in the form of a smartphone application. It doesn’t always have to be that way. It is important to keep it open rather than to define your solution as an “app”. There is a very good chance that the other person has a preconceived notion or a horrible experience with apps in general due to which they have an aversion to apps in general. So if you say I built another app, it might not be received all that well.
Q. How do we look beyond screens and apps completely when designing to solve a problem effectively?
I think a smartphone is a very good solution for a lot of things because it is something that people already have. There is an ecosystem. There is a distribution that’s already in place so it has a lot of advantages to it for sure. So if you build something, people across the world can have it overnight. But it doesn’t always have to be an app-based solution. In such a situation, you start entering into a realm of building actual hardware. And when you do that, you need to be really, really sure that it is something that you are able to scale and are able to distribute it.
Because if you are trying like to build hardware that is not a smartphone based solution, it’s certainly possible. But then the distribution is entirely new. So if you know somebody who wants it tomorrow in Japan, you cannot do it overnight. You need to really scale your business, need to think of supply chain, think of the stakes, etc., so I think that’s like a barrier.
The other advice would be to see if there are other devices that you can build on top of, similar to what Envision did with the glasses. See if you can build your solutions into existing platforms like Alexa or Google Home. I think there are options out there, but it’s something that you have to sit with the user and understand deeply.
Q. Designers today, instinctively try to come up with digital solutions to most problems, but we also know that probably a digital solution is not the right approach for every problem. How do we navigate this situation?
I think software is eating the world. There is not many ways to build really good solutions going forward that only have a hardware component without some accompanying software component as well. Completely non-digital solutions can be built, but to add true value, leveraging software would be indispensable.
I do understand that software is not always the right solution. For example, in terms of how the blind use canes to navigate their surroundings, the level of comfort, input and efficiency of it is not something that can be very easily replaced by software. So, there is a space for hardware solutions as well, but I strongly believe that the best area to explore is the hardware-software combine/integration.
Q. When designing for accessibility, how do you reach out to users keeping in mind any sensitivities they might have due to their circumstances?
I think they are not that sensitive (in my experience) and they get offended when people step on eggshells around them. Everyone wants to be treated as normally as possible. So, if you actually attempt to specifically not offend them, that tends to come across as more offensive.
I was very honest with all of my users since early on about who I was and what I was doing. I found people on places like Facebook, Twitter and other social media. I was very open and honest with them saying that I have no clue how this works, I have zero idea about how it feels to be blind and I don’t think I ever will be able to completely understand but I do want your help so that I can try to.
I’ve realised that it’s okay if you don’t use the right terms as they know that most of the time it’s an honest mistake.
I don’t attempt to make them feel special or speak differently with them. I interact with them as I would with any other person and I think that’s all you need to do as well. There is a very good chance that you will accidentally say or do something that they are not happy with, but as long as you’re honest it should be fine.
Q. Can you tell us about how blind people interact with the Envision app and it’s UI? What are their interactions like ?
So when you’re doing a design for something like a screen reader. It’s the same principles that you would apply for UI design, but you’re to now apply that to just audio.
In visual design, there’s a certain hierarchy, where there’s a certain order in which you want people to notice and interact with the elements. Similarly, for audio, you design in a manner where you guide what they hear first and what follows it and how the flow continues. That’s how you build a map around it.
Q. Do you have any advice for people who want to get into assistive technology?
I don’t think I know too much about how you get into accessible design as a field. I know that it’s something that is being taken a lot more seriously by a lot of tech companies. But in my opinion accessible design should not be a separate field in itself.
I believe that if a company is built the right way, they won’t have a separate role for an accessible a designer because every a designer should be an accessible to designer.
I would ask you to attempt to be a designer for a product or in a space you love and then when you are in there, ensure that everything that you’re doing is accessible.
Q. What were your learnings as a designer turned co-founder?
I’m still learning different things at different stages. I think one of the first things I learnt was, when you’re building something new, especially when you’re just building the app, just push out your current version as soon as possible. In our case, we were never sure what’s the right time to launch it. We had built Envision in October of 2017 and we just kept on improving it in the never-ending pursuit for perfection. But, it was never going to be perfect and finally we just went ahead and did a launch in February of 2018.
I remember we had just put that thing out and within a month we had two thousand people install it and 50 people subscribed to it within a week of the product being launched. In hindsight we feel a little stupid that we could have actually launched it much earlier but we always wanted to improve the product before we push it out.
We need to put out a product as soon as we can so that people can actually start giving you real validation on what is it that has value? If you don’t, you’re stuck in an echo chamber where you’re probably improving stuff that there is no need for you to improve. So, we now attempt to push stuff to the users and start improving them on the fly.
Apart from that, everything comes with a lot more ownership and stake in it because you have skin in the game, right? So you need to really learn to put stuff out that actually has the most value and you need to allocate time very, very well because you can build a hundred things, but we don’t have the time to build everything. So it becomes a big responsibility to pick the right things that will have the most amount of impact for the business.
Another big learning is that as a founder, I won’t get to design as much as I’d want to despite the immense passion I have for building the product. Instead, most of my time will go into building the company and the team that is going to build the product that has been envisioned.
Q. Can you talk about the difference between building a tool and designing a solution?
I think envision oscillates between these two things. A solution is basically that there is a problem and you’re building something that really fits into that problem like a piece of a jigsaw puzzle. The other approach is to build a tool, which is something that users can now utilize to solve other problems.
Envision leans towards a more tool based approach. Mainly because, as it grows, we have more and more use cases that it has been used for, solving multiple problems. Some examples are - some users have been using it to read subtitles from foreign language films, play games like pokemon cards which were earlier inaccessible because they were visual in nature, etc.
So, the approach you take in your design, dictates the manner in which it will be positioned to users and how they would use it.
Karthik was a guest at a community fireside chat hosted for the product design fellowship cohort.
Applications for cohort 2 are being processed on a rolling basis. Apply by July 10th and join the fellowship, free for the first two weeks!Learn More
We publish guides & interviews, organize events, courses, and programs to help you explore and learn design.
Can we keep you posted?