Cambridge Angels backed Signaloid takes the title of being one of the first contract awardees for the Advanced Research + Invention Agency (ARIA). In this interview with founder Phillip Stanley-Marbell, we learn more about Signaloid and Phillip's journey from Bell Labs via Apple to now leading the team at Signaloid.
Can you tell us a bit about yourself?
Yes, certainly, the relevant bits at least! I’m an engineer and I create advanced computing technologies that exploit an understanding of physics. I spent the last 25 years moving back and forth between industry and academia.
In the mid 1990s, while still studying for my undergraduate degree, I spent time working at Bell Labs in the US and got to work under people such as Dennis Ritchie (creator of the C programming language). The experience motivated me to go do a PhD at Carnegie Mellon University (CMU), after which I joined IBM Research in Zurich. After about half a decade there I moved to Apple in Cupertino, where I was involved in the development of the Apple watch, among other things.
While Apple was a lot of fun, I had some ideas at the intersection of computer hardware architecture and probability theory that I knew that, if I didn’t put in the work to turn these ideas into reality, I would regret it at the end of my life. I therefore left Apple for MIT in 2014 with the intention of going into an environment where I could build ideas and a team with the appropriate expertise and then moved to the University of Cambridge as a professor in 2017. I incorporated Signaloid in 2019, self-funding it for two years before raising the first angel investment in 2021 and completing a seed round in January 2023.
And what does Signaloid do?
Many organisations rely on Monte-Carlo (MC) simulations, particularly in the finance and engineering fields. Signaloid’s technology can help them run the MC simulations significantly faster (often twice to ten times faster) than on Intel processors; they get answers faster and have greater capacity to run a variety of simulations. We have essentially designed a new type of processor that enables arithmetic on probability distributions and we can deliver this to clients in a variety of different ways, be it public cloud, private cloud, on prem, or as an edge hardware module.
What are your goals for Signaloid over the next few years?
Our goal over the next few years is to keep growing revenue in our focus market segments of finance and manufacturing and to continue to execute on our roadmap for improved hardware accelerated implementations of our technology.
Our main deployment path for customers today is via our cloud-based deployments but we are on track to complete at least one custom silicon test chip (“tapeout”) before the end of the 2024. We have agreements in place with IMEC and access to TSMC’s advanced technology nodes which will allow us to also provide our technology for integration into accelerators designed by other companies. Interestingly, many of the
more aggressive quantitative finance organizations, both hedge funds and market makers, are now building their own application-specific integrated circuits (ASICs) and technologies like ours that can speed up risk analysis, option pricing, and Monte Carlo methods in general, by orders of magnitude.
At the same time as executing on our medium- and long-term plans, we have current engagements with two of the ten largest banks and we have partnerships with the likes of Bosch and contracts with some of the largest aerospace companies. We’ve also recently won two large UK government agency contracts. We are approaching cash-flow break-even and do not need to fundraise as a necessity, but there are always strategic opportunities (we closed a £3M seed in January 2023 round and have more than that amount in signed contracted revenue over the next two years).
In the context of fundraising and revenue, our goal is to be a commercially-successful entity solving an important problem in the world and delivering value to paying customers. As a company, we judge ourselves by the value and impact we deliver to customers, not by how much we’ve raised in external funding. One area where I would like us to do better, in the daily grind as a company, is improving our balance between large contracts and “sticky” broad-based recurring revenue.
What advice could you give to an aspiring CEO?
Understand how your target customer can use your technology as a tool “in anger” and give them that tool. If you focus on solving your customer’s problems, you will have many opportunities, and real justification, to develop groundbreaking technologies. If you focus on developing a technology, you will likely end up with neither a useful technology nor any customers. So choose wisely.
Before founding Signaloid, I took part in an entrepreneurship programme at MIT where the program drilled into us the importance of Talking to Customers; looking back, while I thought I “got it” then, I think I underestimated how important it is to really understand what each individual customer is trying to achieve. Customers will often initially often fob you off with pleasantries but somehow you have to get them to cross the chasm to trust you enough to share what their real problems are. At Signaloid, we tried to enable customer validation opportunities by always having a customer-facing product, even right at the beginning when we were still bootstrapped off my savings.
A sales person or a “commercial person” that you hire from the outside can, maybe, teach you the ropes of customer acquisition, but a technical founder needs to do this customer acquisition themselves. If you think you can delegate customer acquisition, you are likely abdicating your responsibility. You will get sold many salespeople.
And what advice can you give going into a fundraising round?
If you’re fundraising to be able to say how much you’ve raised, then I’m ill-equipped to say anything that would be relevant or helpful. If you’re fundraising to build long-term
relationships, here are a few positive things I took away from the sustained and elevated pain during Signaloid’s seed round:
- A financing round involves human beings and people are not always rational in their decision-making. This can make it fun! This can make it difficult!
- During our seed round I spoke to at least 60 investors. Of course, the vast majority of these didn’t invest but I learnt a huge amount through the tough questions they asked. I learned a great deal from the experience of all the pitches. I only felt perhaps 2 out of the 60 or so discussions were not so pleasant, but everyone has bad days, no?
- At the completion of our seed round in January 2023, I sent a personal “thank you” email to every investor I spoke with and asked them if I could send them monthly updates. Most said yes. So, since January 2023, I have been sending a monthly update to a list of investors that has now grown to almost 80 (I add new VCs when they get in touch to ask if we’re fundraising). The investors on this list, most of whom did not invest, have been very helpful, making introductions, helping with hiring, and more.
How has having Cambridge Angels in your investor group assisted you?
They have helped enormously and the name says it all; they are a group of experienced angel investors and not just an faceless entity. I have been able to pick up the phone to any of them at any time and they have always been helpful. I am speaking to one of them tomorrow in fact.
You are one of the first recipients of ARIA (Advanced Research + Invention Agency) contracts. Can you tell us about that?
ARIA is quite unlike anything else I have come across in the UK. We have a contract whose award was made public at the end of July and we have other engagements with ARIA in the pipeline. The form of engagements (contract as opposed to grant funding) has many positive knock-on effects: ARIA has several programme directors who are experts in their technology fields and have a clear picture of technology challenges which, if solved, could have significantly positive technological or financial impact for the UK. People interested in the specifics of contracts awarded to date, intellectual property retention by organizations, etc., should visit ARIA’s web page to find out more.
What was the process you went through to get the ARIA contract?
It was an efficient process which took approximately three weeks from the application deadline to the initial selection, followed by discussions with the programme director and their team. The total time from the application deadline to project start was under three months. The ARIA programme directors are of such high calibre that, even if a contract tender is not successful, you will likely find the process intellectually stimulating.
