It’s been a five-year journey, but officially Yuty was incorporated last year
So 2015 – which was pivotal for AI, we’re talking about driverless cars, manual labour being automated, and was also the year that I fell pregnant, and the year that I was interviewed by a beauty journalist. She said, ‘Simi, you know, your skincare’s amazing. You’ve shaved your hair, you’re rocking this style.’ And I said ‘It seems super glamorous, but actually I shaved my hair because I was frustrated.’
I was having to trek 3 hours to get my hair ‘did’ every single week. I’d rocked weaves, extensions, various styles, and I just wanted to embrace my natural hair, and I just wasn’t able to find the right products.
She asked me about my skincare, and I said, well, I’ve got combination-oily skin – I didn’t realise at the time that I was pregnant – and I was just like, something’s not quite working (laughs). What used to work for me isn’t quite working. My makeup is doing this strange thing. I’d basically given up.
At the time I was at a business where I was matching PRs and brands with celebrities and influencers to help them drive press exposure, and I was thinking, what if I could just match people with beauty products using AI?
I started to teach myself AI…
It began with teaching myself Python, Ruby on Rails, getting to understand quite specific coding languages, looking at toolkits and libraries and seeing what was available to me. And I was realising that these libraries, and the data available – they’re kind of biased.
I’m Black, and I was like, where are the people that look like me in this? I’m trying to figure out this algorithm to match people to products, and I can’t quite find the right data to even test if this algorithm works. It was lots of crowdsourcing of data, asking friends and family, understanding different skin conditions.
After about nearly 4 years, I thought, right, it’s time to do this for real
I had the golden handcuffs, my salary was incredible. It stops you from really going after it, because you don’t want to disrupt the family. You understand that when you undertake this entrepreneurial journey, it’s going to be scary.
So I did try in 2019, realised I needed more money, went back to work, and saved for a further six months. I had my data science and machine learning lead alongside me while I was doing this, I was studying AI at Oxford, saving money – all of this is happening alongside having children.
August 25th of 2020, we launched our MVP. The rest is history.
Machine learning is data hungry. It won’t work without enough good-quality data. I’m so cognizant of the fact that people look so different – it’s really important that data reflects what’s going on in society, so that we can make those accurate predictions, otherwise we’re failing.