The Story of an A.I. Adventure into Fashion and Fit.

The Fit Problem: A $trillion problem and The Space Race for Fashion.


 I started MAN London because most t-shirts were too long for me, I had a problem, I thought other people would too… so I thought I had solved it by designing my own ranges of clothes for a shorter torso - simple.


This is where I would end up… but this isn’t where I started.

The Ultimate Vision… Clothie in the Cloud no app just tech that plugs - Facebook Messenger, WhatsApp etc etc…

The Ultimate Vision… Clothie in the Cloud no app just tech that plugs - Facebook Messenger, WhatsApp etc etc…

Download the deck… view the    website here

Download the deck… view the website here

Screenshot 2019-07-15 at 11.30.08.png

Standing in the MAN London store I would watch in horror as men tried on clothes I had made to fit me… which didn’t fit them and looked horrendous.

I mean people with my body shape 5’9” they looked great on but…

..the writing was on the wall that first day when Jeff Goldblum walked in - all 6’6” of him errr not sure anything in here will fit you Sir!


When you make clothes, during the design phase you get a spreadsheet from China that is the fabric cutting sheet - the actual measurements they use to cut the fabric… that they plug into a CAD machine and that are very detailed

Clothes start life like this:



It struck me after MAN that all this data of measurements is deleted and replaced with an S… or an M or a L or an XL…

and that I have ‘designed’ my M to be a certain set of measurements…

..other brands have chosen their own measurements for a medium etc etc and all the mediums across all brands can therefore be different.

Then the consumer chooses S or M or L or XL.. all of which can be different measurements - which is mad and really confusing.

Why would you delete all this data and replace it with S, M, L?


Why not measure peoples bodies and match that data to the brands’ garment data

the consumer can then JUST shop the clothes that WILL fit THEM…


iSizeme was Born


iSizeMe was born, albeit briefly.

I went to try and work out how to measure peoples bodies easily and accurately (not so easy)…

..using their phones and having pitched it to scientists from Imperial College London we worked on a solution:

Just take 2 images of your body and we could measure you accurately as long as there was something in the image that we new the size of... sheet of paper or sticker or something.


The fact you need to do it in your underwear was a worry and try and position the camera away from you, but still get all of you - is really not that great but to solve the $1 trillion fit problem, to save you ever having to try anything on again? Surely that’s worth it?



Obviously I rushed out to patent it…

I was so worried someone else might develop the tech or patent it before me - if this is you… my advice would be - do searches, use the searches as research and then think does this really need a patent?




The market today is so fast moving that a patent can actually prevent you from changing and moving fast enough… the shear cost and complexity for a lean start-up could kill you.

If I had the £25,000 spent on a patent, that we never actually were awarded - I could have used it later to make a complete working prototype MVP - The A.I. and the cloud piece all joined up.



Armed with a patent pending I produced an investor deck, then another one, errr and another one, a million decks later, pitch deck, pitch deck, pitch deck and zero meetings or interest… which judging by the deck below is not massively surprising - it’s terrible!!


Einstein once said doing the same thing and expecting a different result is the definition or insanity… he was right!

But I thought… well the decks ARE different ;-) So I made more and more… and more.

Today with a million digital businesses to chose from investors WANT data, WANT a large market, WANT proof of product market fit, WANT evidence of revenues - because other people have it… so if you don’t then lets hope you went to Stanford ;-)



BUT we finally landed a meeting with the Head of Marketing at ASOS… they have a massive returns problem - the meeting was arranged through a friend of a friend - i.e. not from one of my decks 😩!


we pitch him and within 5 minutes he ripped the idea apart - very effectively.


“Peoples body measurements don’t determine their fit preference.”

“2 people with the same bodies might prefer very different fit and he would know.”

Game Over. 

Game Over. 

Game Over. 

Game Over. 


“Oh and by the way we (ASOS) have no clue as to what is actually in the plastic packs from China.”

Companies like M&S employ people to measure clothes as they come off the production line to make sure they’re the correct measurements but at ASOS price point this isn’t economical.

Game Over. Game Over. Game Over.

Game Over. Game Over. Game Over.

Game Over. Game Over. Game Over.

Game Over. Game Over. Game Over.

So I left dejected, my iSizeMe in tatters…

You really can not ignore someone like that, telling you / saving you from yourself.

and then I had a brainwave! 


Instead of measuring peoples bodies which is pretty icky anyway…

Just measure the clothes you already own,

the ones you wear all the time,

use those measurements to match you,

to data of new clothes…

snapfits become Clothie - Selfies for Clothes!

Screenshot 2019-07-15 at 11.34.37.png

pronounced CLOTHE V NOT CLOTH V - everyone calls it cloth V

- Note to self - I really need to stop making words up -


This is much easier… no more bodies to measure, no stripping off, just flat 2D patterns to measure.

Now just one picture rather than 2.

A picture of your favourite t-shirt opens a whole world of t-shirts to you… BIG VISION!


  • Omnichannel

  • Personalised

  • Unicorn business model..


I’m excited… I do a ton of research in measuring apps, I find a Jon (amazingly talented) in America who is using A.I. Machine Vision to measure buildings.

Dear John, Could you measure clothes with your technology,

Dear Dan, Yes I can….

the clothie prototype is born.


The budget is minuscule, I wireframe and design and the UX myself to save money

and find a designer from 99designs to polish it off / try and make it look a little more professional.


3 months and £5,000 later BOOM I have a prototype app, I have a WORKING A.I. prototype!

Jon is a genius…


I have a new pitch deck - I AM Excited!

This is the deck that attracted Huddle - a London digital agency (an emoji phase):



I designed a rail of t-shirts with my pitch printed on…

this also helped me land the London agency Huddle who wanted to get involved, in return for equity.

The prototype at least works, for a first go at A.I. and for just £5,000 it’s pretty impressive really - VERY accurate.

But you do need to follow the instructions, there are way to many steps, the measurements it was pulling off didn’t match the measurements in the manufacturers tech packs and the A5 sheet of paper is not great UX.

BUT this is a prototype and the funding was going to correct all of those issues.

It is however much better than stripping off and measuring your body which, as we now know still wouldn’t find you clothes that fit you…

The Science Bit - How the A.I. Worked


The prototype uses an A5 sheet of paper to create and accurate scale - which on new phones you wouldn’t need, hence why patents can be an issue because if the method includes A5 paper it’s already put of date.

The A.I. finds the outline of the paper and the outline of the garment, deletes the background, changes the perspective of the garment and measures it!

Clothie Deck - HUDDLE.001.png

The agency design us a new brand and a new deck! We spent time on the actual process too with Danny and Tristan and turned the app idea into a SaaS / Cloud process model - amazing… so you take one picture of your favourite dress and next thing you know you’re getting dresses that fit you straight to your Facebook Messenger or wherever - I mean what is not to like.


The Clothie Cloud Could Revolutionise Fashion FIT First.

Clothie Cloud V3.001.jpeg

Imagine only seeing or shopping clothes that actually fit you…

Clothie Cloud V3.002.jpeg

BUT (ignore everything before a BUT) with our lack of resources we can’t realistically get the data we need to pitch this to a VC or investor.. the investment world has changed dramatically since VizWoz and you need your pitch to be evidence based not napkin based - unless I assume you’re MIT grad with a genius coding brain.

We did well with the resources we had…



Also in 2015 the high street is showing the pain but not anywhere near enough for the brands to look for a solution… it needs to get worse still before it gets better scenario - So we’ve parked the technology, it’s sitting here if you want it and I moved out of fashtech.

I thought the retail landscape would rapidly change and it is, brands on the high street are in real trouble, no high street, no changing rooms and at some point the brands are going to need to come together to solve that problem - the brands USP is that they do know what’s in the packets because they have people checking at the end of the production line and maybe just maybe they might want to encourage customers to take pictures of their clothes that fit them the best… only time will tell.


I have a potential opportunity to go back to EdTech which is where it all began so I make the move…