“Our spin-off will follow in Logitech's footsteps”

Amin Shokrollahi, founder and CEO of Kandou © 2020 Alain Herzog

Amin Shokrollahi, founder and CEO of Kandou © 2020 Alain Herzog

On 17 November, Kandou Bus, an EPFL spin-off, announced that it had raised CHF 92 million in a Series C funding round. This brings the total funds the company has raised since 2011 to over CHF 130 million, placing it in the top five EPFL spin-offs by this measure. Its sophisticated computer chips form an integral part of the race to boost data transfer speed while lowering power consumption. Amin Shokrollahi, the company’s founder and CEO, is convinced that their technology could potentially be used in all types of communication systems, including satellites, within the next decade or so. We spoke with Shokrollahi about his vision for Kandou Bus.

Despite the pandemic and the ensuing economic fallout, your company managed to raise CHF 92 million in funding – CHF 56 million in 2019 and the rest very recently. What's your secret?

It was precisely because of the pandemic that we decided to streamline our fundraising efforts by continuing a third round rather than starting a new one. So we signed up the same set of investors, mainly Swiss, as well as a few others they had brought on board. This eliminated the need to make lots of presentations to attract new investors, which would have been difficult in light of the public-health situation.

In 2011, when you were a professor at EPFL, your lab developed an algorithm that transforms how computer buses transfer data. This breakthrough technology was the subject of your first patent. Is that when you came up with the idea for a start-up, with the anticipation that it would take off?

Within an hour of our system working for the first time, we thought it would be a good idea to create a start-up. It was immediately clear that, while our technology could give rise to some outstanding research papers, the best way to spotlight it would be through a company. However, Kandou Bus grew much more slowly than we had anticipated in our original business plan, which called for a few months’ R&D work at EPFL, a prototype in under a year, and funding rounds the following year to turn the system into a marketable product. We were hoping to get where we are now in just four years. But when the applications for your invention are in brand-new territory, as was the case for us, then the process is like skiing down a treacherous slope at night with only a flashlight. You have limited visibility, you have to react with lightning speed and, if necessary, change course without falling over. Under such circumstances, finding your way takes time.

What were the key development steps for your business?

We're lucky that our technology is a horizontal one – that is, it concerns all segments of the computer-component market instead of just one narrow application. The downside of this was how to gauge the market early on to determine which segment to target first. During this phase, which lasted three years, we were in stealth mode; we didn't share what we were working on. Fortunately, the CEO of a US start-up, which had become a multinational some years earlier, believed in our technology. He loaned us CHF 10 million in 2012 so that we could keep our head above water. Then in 2014, after delivering a paper at a high-profile industry conference, someone from Marvell, a semiconductor company, expressed real interest in our system. It took two more years to secure this first contract, which really set us on our way.

Successful IT start-ups are rare in Switzerland. What are the advantages of being based at EPFL?

First of all, I live here and I love this region – that's important. Also, because there's not much ICT here, it's easier to blaze your own trail. At a place like Silicon Valley with so many successful high-tech firms, newcomers have a tendency to imitate them in order to get backing, even though that might not be the best course for their business. Finally, being positioned outside the microcosm, and its attendant discussions, lends us an air of mystery, which is an advantage during negotiations. Locating our company at EPFL was a strategic choice at the beginning. I was a professor there and needed to be able to shuttle quickly between the company and my office. Now, however, it gives us a clear advantage in hiring talented young engineers and keeping abreast of research developments.

What advice would you give to entrepreneurs want to create a fast-growing start-up?

Being an entrepreneur is a lot like being a scientist – you have to be passionate about what you do from the outset. You can't invent it later on. Without that kind of enthusiasm, it's extremely difficult to succeed because your personal investment is crucial: developing ideas takes up a lot of time and you don’t get much vacation or weekends off. This kind of commitment is also what allows you to weather the storm when nothing goes according to plan. Afterwards, the fast growth comes on its own. There's no big mystery about the early funding period. You have to cultivate your network, but if an investor doesn't latch on, there's no point spending huge amounts of time trying to convince them. You can see right away if someone's interested.

What are your plans after this substantial funding round?

Now our main goal is to market our first chip for all USB-C connections by spring 2021. This chip allows for faster, lower-power data transfer between all types of devices: smartphones, computers, external hard drives, and so on. We’re currently in the final development phase for the high-end version. Over the next year, we’ll roll out other chips aimed at a much larger market. This will be our first widely-distributed product.

Where do you see your company in ten years or so – is an IPO in the cards? Will Kandou be the next Logitech?

In terms of success, I think Kandou Bus will follow in Logitech's footsteps. My goal is to supply chips for all the devices people use to communicate. Today, systems are built using components, and we’re designing those components. But we’re planning to fabricate finished products based on our technology within a decade. Our technology will likely be used in satellites. With regards to our business model, we want to remain an independent company. As a CEO, I have to keep an eye on the ROI for investors and future shareholders, but as a founder, I have a strong link with the company and want to remain with it for as long as possible. An IPO would be the best way to combine the two.