The Shazam for clothes finds any outfit, even this one, then lets you recreate it using cheaper brands

Interested in clothing and style? Then firstly, what are you doing reading a tech blog? But secondly, we may have found the ultimate system for buying new clothes, without having to venture to shops.

Just as Shazam miraculously recognises tunes from being played snippets of them through dodgy shop PA systems, startup Orpiva reckons its app can recognise clothing in outfits it's shown photos of, whether from the web, magazines or random strangers you've photographed because you like their dress sense. Bit creepy.

The claimed cleverness does not end there. You can also browse and pick clothing from 15,000+ retailers and see it mapped onto the contours of your body - you just need to provide your height, weight and a front on photo - like a paper doll.

Because it's finding clothes strictly through machine-learning image recognition, rather than by a bunch of guys in India or Scotland tagging clothes as “red”, “sweater”, “cotton” in 10-hour shifts, Orpiva should be highly accurate. Search for a red shirt (hey, you might!) and it'll actually find shirts that are red - try that with Asos - as well as colour combinations of your choice.

You can then “try on” outfits virtually, which should reduce the level of returns. As well as e-commerce, if you're out and about, the app can also show nearby shops where you can try on and buy the garment of your dreams.

Perhaps cleverest of all, Orpiva is also said to be the ultimate “get the look” app. Show it a photo of an outfit you like - this will need to be fairly front-on and static, due to the limits of image recognition - or find its component parts from Orpiva's database, if you know what they are. Then, you can scale the cost up and down. So you start with the Armani version and end up with the closest match that the likes of H&M can provide - Orpiva has cut deals with retailers from high end to high street.

If you're shopping to suit your own look rather than Paul Weller's or David Gandy's, you can also search by garment types, throw together virtual outfits from multiple brands and see them on your avatar, then share with friends to find out if, for instance, your bum looks big in them. Orpiva will select from the mammoth selection of clothes on offer based on a basic taste questionnaire and your previous buying history.

I see new stuff all the time, and my usual reaction is “Meh”. But this is genuinely exciting. If it doesn't measure up to the claims made at today's launch, I will be mad as hell. However, I am cautiously optimistic that it will measure up (ho ho). That's because the people behind it not only physically resemble the cast of Silicon Valley, they also have serious pedigree, having worked on computer systems for everything from robots to keyhole surgery, with seasoned “Expert Advisor” George Thaw having been COO at SAP and head of product services at Microsoft.

Because they're tech guys not style guys, they're looking at fashion with completely fresh eyes, and what they've come up with sounds amazing, if you're into clothes.

The only possible problem is that compared to mastering the fashion world and keeping abreast of changing styles, it may turn out that making robots, brain surgery and running global corporations are all actually pretty straightforward. Fashion, after all, is danger…

You can check out Orpiva here.