Ahmad Moussa, a former executive at Dubai telecom du, may not be a fashionista but he is keen to make his Stylar app, the Shazam of fashion with its ability to parse images and match the colour, taste and features of an outfit spotted on a celebrity within a few seconds.
“The fashion discovery starts with an image,” says Mr Moussa, 41, an Egyptian-Canadian who co-founded Stylar and is its chief technology officer. “People go on Instagram or Google images, they look at celebrities, what they’re wearing, their style, look and so on, then they end up with an image that has clothing items in it that they want to find at their price range.”
That is where Stylar, set up in 2017 in Dubai, comes in. Mr Moussa created the app, which uses images that will leverage artificial intelligence and monetise it. With a team of 10 full-time employees dispersed around the UAE, Poland and Singapore, he intends to grow the company to compete internationally and not be restricted to the region.
As fashion is one the fastest-growing segments of the burgeoning e-commerce industry, customising the technology to the appetite of online shoppers to buy the latest fashionable outfits at a suitable price was a no-brainer. The app will have special features that can narrow down an image to a 2,500 colour palette once it is augmented in a few weeks.
“We curate some images to have some sort of catalogue for looks in general, but the typical user case is that you already have an image and that you want to buy this kind of look or outfit, and you just don’t know where to buy it from, or you want to buy it at your price range,” says Mr Moussa.
“We will identify the item, we will recommend similar items, then you can filter by price range, retailer and designer. And then you can get a similar item that you liked, that is similar to the one in the look.”
Working in fashion AI is not just born out of a whim to dabble in imagery. Mr Moussa graduated with a master's degree in management analytics for AI in 2015 and has a passion for computer vision and image analysis using the technology. E-commerce figures may prove his value proposition is on the mark.
Fashion products are purchased over the internet by one-half of all online shoppers around the globe, according to a report from ResearchAndMarkets.
By 2020, fashion e-commerce is expected to net over half a trillion euros (Dh2tn), with online penetration for total retail sales exceeding one-third of all clothing, accessories and footwear purchases.
The growth is sparked by a sharp rise in the number of shoppers online “who look for convenience, larger product selection, lower prices, convenient payment methods, online reviews and product ratings when purchasing clothing and accessories on the internet,” the report said.
Mr Moussa says: "If you write black dress you will end up looking through 1,000 images to get to something that is similar to what you want and then you have to compare the price and so on.
“We do an automatic reverse search on all the retailers’ catalogues, and we present you with the results that you want. We basically narrow down one hour of search effort into a few seconds.”
The main ingredient for his app, which is available on IOS and Android stores, will be his AI proprietary technology that will augment the app allowing users to upload an image and get recommendations on it.
“AI is the voodoo for computer science for a reason,” says Mr Moussa. “AI is a very big wave. It is the internet starting all over again.”
Mr Moussa expects to sign a contract with a big regional e-commerce company in the next few months, helping him target the business-to-business segment of the market.
“We have a very good team. We understand AI and we understand investment opportunities. We are not going to be cheaper,” he says.
His focus will remain on the business-to-consumer market, which he expects to generate 70 per cent of the revenue.
The app tracks around 35 retailers and he expects the number to go up to 100 by the end of the year.
“For the time being, I am just tracking the wider retailers with the deepest catalogues so that I can provide more options without having to do a lot of effort,” Mr Moussa says.
"That's why I go after Amazon, Nordstrom, Macy's, all these big players around the world that have multi-designer or multi-label brands under one e-commerce shop."
He projects that his active user base will range between 30,000 and 50,000 by the end of the year. The company makes its money from affiliate commissions from retailers it recommends.
“Every time you go through a link from an affiliate to the retailer, they drop a cookie. This cookie is active for 30 days and then anything that you purchase within 30 days, we get a commission on it. The commission ranges from 7 to 10 per cent across the board in all the international markets,” he says.
To reach his targets, Mr Moussa is looking to extend an angel funding round with family and friends who have already invested $250,000. That is in addition to the $500,000 put forward by him and his co-founder Basil Moftah.
Mr Moussa is looking to raise another $300,000 to $400,000 from that angel round to help enhance technology and expand Stylar's marketing budget, particularly to build up the user base in the UAE and Saudi Arabia.
At the same time, he is talking to venture capital companies in the two countries as well as internationally in China and the US, to help with the next phase of growth after the angel round ends in two to three months.
Securing VC money, however, is not easy.
“I am being treated like another start-up that just started an e-commerce site. Investors with the right mind-set, they would understand that they are investing in a technology capability that is beyond what is available right now,” says Mr Moussa.
“It requires deep pockets because I am competing with the titans and I need the muscles.”