My name is Appu Shaji. I come from a technical background and computer vision is a domain that has fascinated me for the last 25-30 years, starting from my late teenage years. I hold a PhD and Post-Doc in computer vision and worked on several well-cited academic papers which can be found in popular computer vision libraries around the world.

Back in 2012, while I was doing my Post-Doc at EPFL, Lausanne, I took the plunge into entrepreneurship by starting my first company which sadly failed. The second one I started, I sold to a Berlin-based company called EyeEm and Mobius Labs is the third company I formed.

What inspired you to launch your business and what is the end goal?

As someone who has been in computer vision entrepreneurship for a long time, I can confidently say this is the most exciting time for this technology. Several extremely critical applications that would promote revenue growth and cut down operational costs are going to use computer vision technology which is being widely explored in several industries. But in order to set this up, businesses still usually require a huge investment and effort to find the right talent, the right data-set, and the efficient training of computer vision models. To make all these processes simple and accessible, we at Mobius Labs sell software that anyone in the world can use in their applications without the need for any AI expertise. Our software has two parts: the first is a built-in algorithm that does image and video analyses and additionally, has the ability for end-users to train these models using a no-code AI. So instead of having to go through a tech team, anyone, starting from a product manager to a designer, can train these AI models. 

How do you set yourself apart from other businesses in your industry?

Typically, AI companies build their product with a developer crowd in mind. In contrast, we are building our software to address non-technical users. This is important on many levels. For instance, a media organisation that is sitting on millions of images and wants to build a next-generation visual search automation tool, must make a number of decisions in order to successfully adopt AI systems. These decisions are best made by the people who understand their users and know their business well: product and business owners for example. So, making a computer vision software that is accessible to them will allow better products and applications to come in. By targeting a wider audience, we improve accessibility and use of this technology.

What’s the biggest lesson you’ve learnt so far as an entrepreneur?

It’s probably a bit of a cliche, but following your passion is actually one of the major ingredients of entrepreneurship. But what needs to be understood is that it is not the only ingredient. In my journey as an entrepreneur, there were a lot of times that were tough but having some sort of internal clarity as to why you’re doing what you’re doing really helps to get through these challenging moments. It really helps this whole journey and is significantly important to keep going back and re-evaluating this. There are two cases where your entrepreneurial journey can get stuck. First is when you run out of steam: the things that attracted you when you just started out might not later on and can cause you to lose motivation. Secondly, it's when you get too comfortable where you are, staying inside your comfort zone and don’t push yourself enough.  It's about finding the right balance between selecting a tough enough problem that you have a good shot at solving and an interesting enough problem that you are genuinely curious to solve, and then being one of those few people who are actually motivated enough to go and solve it. I think there are only a handful of people in the world who can achieve this.

With all the success stories around entrepreneurship and how innovative people have to be to take the leap. How do you think you’ve innovated your sector and why?

Building an organisation that takes computer vision to a larger audience has always been my biggest motivation for a long time. I have had partial failures and successes along the way. The first company that I formed in 2012 was also related to computer vision technology. Back then, I had a purely technical outlook on life, which stemmed from the fact that I was immersed in academicia most of the time. I didn’t realise that working in this industry is a totally different ball game altogether. A lot of mistakes were made, but it eventually turned out to be a great learning experience which helped to innovate and look at things from a different perspective when I formed Mobius Labs.

What are your thoughts on failure? 

Failures are an essential part of entrepreneurship. This is especially true when you’re a tech founder. Technology has a different notion of failure in the sense that when you’re an engineer who’s building a solution, it should be one that should essentially not fail because you don’t want to build solutions that don’t work. The classical training that you get in universities is that you should do something in a way that is as precise as possible, so the chances of failure are minimised.

However, this might not always be the correct way to go when you’re attempting to do something like building a company because here, you’re trying to solve a hypothesis that might end up in failure. Instead, you just go from one failure to another, and you’ll eventually hit the place where you want to be. During the first few years when I switched from academia to entrepreneurship, I was really overthinking and over-building things. But through these failures that I’ve had, I now instinctively think about what the simplest solution to a problem is (as opposed to the most fail-safe one), validate the solution, and then work on raising its quality. Not getting afraid of failure is something I had to learn the hard way and it’s there in the back of my mind all the time.

As a business owner, do you know when to walk away from a sale? 

For any sale, it is a two-way relationship, and a successful sale should be a win-win for both parties. There are cases where you see this isn't the case, for instance, with a very big client, the cost of doing business, which includes legal costs, negotiation costs or cost of custom projects that we have to do for them, might be too high. In that case, we should weigh up the opportunity and see if this would turn out to be a lose-win situation for us. At the other end of the spectrum, a company might want to buy our software not as a necessity for their processes but simply to try out the technology. In that case, the probability of the deal closing is pretty low. If we can’t, as a company, decide what sort of value we can bring to the potential customer, it is better to disqualify that lead. This can prove to be a good product learning experience because ultimately, if you can build a software that can find use-cases for multiple companies and can be resold many times over, you win as a company. Sometimes when you try to convince people to buy your software even though the value-add is not explicit, you waste both your own and your customer’s time. Here it is better to walk away, and instead use it as a learning experience and come back after a few months with a better feature.

What plans do you have for Mobias Labs over the next two years?

At Mobius Labs, we have built an extremely powerful technology which is able to take computer vision technology to a lot of applications. As a scaling company, to take that product to a go-to-market strategy, we explore one vertical at a time.

We’ve learnt a lot by working with multiple companies in the media sector like press agencies, fashion companies and stock photo platforms. Now we will be adding multiple verticals in our business units as we scale up further. One vertical that we are thrilled and passionate to explore is the space sector where we can take our very light solution to satellites. Several extremely expensive satellites are being launched into outer space; these satellites, which are essentially high-resolution cameras taking images of the world, would suddenly have the ability to get smart using AI. This is something that is really exciting for us. We’re also looking at other sectors like manufacturing.

How important is company culture and what is your top tip to get it right?

Company culture is paramount, especially when it comes to the software sector because ultimately this is technology that is being used by millions but built by a small set of people. Everyone’s contribution is really important. In general, making the company move forward as a single unit, in which all the members are aware of what the company stands for and what everyone else is doing, fosters a strong company culture. This way people can look out for each other when it comes to it. Building a unit where you have teammates with strong individual skills and, at the same, have a good mutual understanding of what our priorities and values as a company are and where we want to go is the key to getting our company culture right. It’s an extremely interesting and challenging aspect for us and at Mobius Labs we are now 28 individuals with strong skills, opinions and backgrounds. The key is to find a common thread that helps us move together and supplement each other; finding that magic elixir that everyone is looking for! Once you have a team that supports each other, your job as a CEO or manager is done. Getting that right team structure and attitude in place is very critical and is a process that needs to be worked upon and improved constantly. I’m very grateful to have that at Mobius Labs.

What’s the most important question entrepreneurs should be asking themselves?

I think one of the main questions you ask yourself as an entrepreneur is what value you are bringing into the market or society, how that value is relevant, and if it is relevant then how you can do it at scale.

How do you believe the evolution of tech will affect your industry over the next 10 years?

In the next ten years, everything will change - that’s the short answer! But before going into the next ten years, I would go back ten years and look at what has transpired. This was probably the time that the first iPhone launched, and computers spread to all businesses and households. The exact same thing is happening with AI now. Computer vision solutions especially are at its inflection point; every year you’ll find a better solution, a software that is able to understand speech, vision, language in a much more efficient way. This is why we name our application “Superhuman Vision for every application”. Superhuman capabilities in every aspect: whether it is to detect objects in millions of images in a few seconds, or detecting a specific object with high precision will become easier in the next ten years, and we hope to be the company that provides the solution for that.