AMA: Michel Valstar, founder of BlueSkeye AI

Founder, Chief Evangelist and Scientific Officer Michel Valstar and the team at BlueSkeye AI are looking to save lives with their revolutionary AI facial recognition app. 

Content warning: mental health, suicide.

Hi Michel! Could you share the story behind your startup, including the problem you set out to solve?

About one in five women suffer from poor perinatal mental health, 70% will hide or underplay their illness and suicide is the leading cause of direct maternal death within a year of having a baby.

Added to that, an estimated 8 million people in England with mental health problems cannot get specialist help, and the cost to the NHS is about £1.2 billion a year. On top of that, the cost to society is a further £8 billion.

It doesn’t have to be that way. Our face and voice AI can accurately determine your facial muscle actions, gaze direction, and tone of voice.

By getting mums to undertake simple and fun tasks regularly, such as reading a story to their baby or completing a pregnancy diary on their own phone, we can detect changes over time in this expressive behaviour that can indicate problems with mental health.

Mums-to-be can then share that information with their doctors giving them frequent, accurate and comparable measures of their mental state.

We believe our technology will save lives.

In 2024, we will start running a clinical trial of the TrueBlue app with the NHS and the Institute of Mental Health to establish that the app is safe to use by women during and immediately after pregnancy.

The trial will assess how acceptable and usable the app is in real life, and how closely the AI mood assessment agrees with standard clinical measures of depression.

Incredible. Best of luck with the trials! What sets your AI technology apart?

We are very good at objectively and repeatedly measuring medically relevant face and voice behaviour with AI, using readily available and accessible technology e.g. mobile phones. We believe we are the only company in this space that uses face AND voice. 

We continually assess someone’s expressed emotions. Our proprietary technology, which has been trained on millions of items of data, means we can distinguish between a large number of emotions, dynamically highlighting changes in someone’s feelings and indicating their intensity.

This provides quantifiable, repeatable, and more objective measures for things that have traditionally been hard to measure and have often been measured subjectively. We have the capability to go beyond quantifying emotion to measuring other medical conditions that show up in the face, for example, Parkinson’s.  

Our technology is developed to the highest clinical standards and is based on nearly 20 years of research, published in over 100 peer-reviewed papers, and cited over 17,000 times. 

Did you hit any roadblocks along the way?

We hit many barriers great and small. Perhaps the most significant was the amount of time it took to raise seed funding. This coincided with the slowdown of the global economy and investment was drying up worldwide.

It took a lot of my time to raise our seed capital. This meant I couldn’t spend that time building up the company and accelerating the development of our products. 

Have there been any significant changes or evolutions in your business strategy?

We changed quite early on when we decided to focus on medical applications, something we hadn't done for the first year.

The second big change resulted from going through a limited-time market analysis. We picked six big, interesting markets and we took those to CES to see what interest there was. It turned out automotive manufacturers saw a huge need for our skills and we left CES with a client.

And now we are undergoing a third change with winding down the amount of customisation and proof of concept work we do and moving more to a subscription and licence-based model. 

What's your view on mentors? And what advice would you offer to others seeking mentorship?

I’ve got a strong view on this - even before I started out as an entrepreneur. In my view, mentors are to be sought out by the mentee not forced on them and they should regularly change.

You should find mentors to help you with what currently is your greatest need. That means changing them over time.

I’ve had a balance of formal and informal mentorship at BlueSkeye with three or four different mentors over time. I joined a CEO “tribe” at some point which was really helpful for about 18 months. It showed me how other CEOs operated. I had a mentor who helped me with some very specific operational issues at some point as well as mentors who helped with the more strategic aspects.

As a first-time founder you have to learn an incredible range of things very quickly and if you don’t have a good support network you can come unstuck!

Let's talk about your team. Can you discuss your approach to assembling and motivating folk and the role they play in your company's growth?

We focus strongly on culture. We want the culture of the company to be mission-driven, to build something that will do social good. We also set out to be mavericks by setting up outside the Golden Triangle, establishing ourselves in the East Midlands and capitalising on local talent.

For initial recruitment, we relied on our personal network. I raided my old lab at the university and recruited people from there as we grew from the university more widely.

We’ve built on our developing reputation as an interesting place to work, where people can devote their time to working on challenging and socially meaningful projects.

I’m very pleased to say that we always get many very high-quality applicants when we advertise a post. As a result, we can often avoid using recruiters, in particular for our technical roles.

Selecting the right staff is very important. We found it easy for tech, because the three founders are experts in their fields. For commercial appointments, we relied more on our advisers. We were lucky to sign up three very experienced advisers who were very generous with their time early on. 

Do you have a share scheme in place? 

We have a couple of share incentive schemes; one is an Enterprise Management Incentive (EMI) scheme, primarily to give everyone in the team a sense of shared ownership and shared goals.

They are all going to reap the rewards of our eventual success.

It also served as a way of remunerating people in the early days when we were paying below-market rates, and that attracted talent.

We also have a growth share scheme for our advisors so we can reward them for their help and their networks. They couldn’t take advantage of an EMI scheme as you have to work more formally for the company. 

Brilliant. What's the most unexpected or quirky thing you've learned or experienced as a startup founder?

One of the quirkiest things is a lot of what you see in the show Silicon Valley is true! The ramshackle offices, the coffee machines, the strange meetings in strange places, especially in the beginning when you’re a small team and you don’t specialise. We’re a bit bigger and a bit more normal!

If you were stranded on a desert island, which of your skills do you think would be most helpful?

I’m a lumberjack and a certified scout, so I reckon my chances would be better than most.

We know who we’d want around in an emergency! Thanks, Michel!