Leon Szeli is the CEO and Co-Founder of Presize, a fashion-tech startup using AI and Computer Vision to provide the most accurate body measurements from a video and/or body shape questions.
Before Presize, Leon worked in HCI research at Stanford and Cambridge. He holds a B.A. from LMU, a M.Sc. from TUM, and an honors degree from CDTM.
1. What is Presize?
Presize is a Munich-based startup that helps fashion online stores to increase their conversion rate and reduce their return rate. We help the customers of s.Oliver and Co. to choose the right size of clothes with the help of a smartphone video or by answering questions about body size.
We offer an add-on for existing online stores and our service is free of charge for the end-user. Actually, the user doesn’t even notice that he is using our service, because we appear in the branding of the store.
The stores that use our technology pay a one-time integration fee and then have a three month trial period to test our product. After the trial period, customers can choose between a fixed monthly fee and a variable commission that is generated when a customer purchases a garment whose size has been determined by Presize.
How does Presize work?
If you want to buy clothes online, you are asked general questions about weight and height. Afterward, you have the choice to record a video of yourself with your own smartphone. To do this, you place the smartphone on the floor and lean it against the wall. The background in the video doesn’t matter, you only have to rotate 360 degrees once. It also doesn’t matter which clothes you are wearing during the scan as long as you can see the body shape properly.
The app then creates a 3-d model of you and extracts body measurements from it. These body measurements are then automatically matched to the size tables of the clothes.
However, if you are not able to record a video, because you are currently sitting in the subway, you can skip the recording step and answer a few questions about your body shape.
As soon as the customer uses our service, buys the garment, and returns or keeps it, the algorithm learns how users with similar body measurements fit the product and therefore improves its accuracy. This is the AI component of our product.
2. How did you come up with the idea for Presize?
My co-founders and I met as students at CDTM. Back then, we had a university project with Hoffmann, the European market leader in B2B workwear and tools.
When a company orders work clothes for its employees from Hoffmann, the body measurements of the employees are first taken manually and the different pieces of clothing are tried on.
We wanted to solve this extremely manual work and built a first prototype that automatically recognizes body measurements and it worked quite well. We then wanted to pursue this technology further and applied for a patent for it. We thought that, especially in eCommerce, the right size of clothes plays a crucial role since the return rate is on average 50% and many online stores don’t let you put different sizes of the same product into the shopping cart anymore.
Can you tell us more about yourself and the co-founders?
I’m Leon, the CEO of Presize and I take care of sales, legal, finance, recruiting, marketing, and investor relations. Previously, I did research on technology innovation at Stanford and founded my own start-up for B2B video streaming. I then studied Consumer Science at TUM and did research on user experience in AI at Cambridge.
My co-founder and CTO Tomislav is originally from Bulgaria and studied Artificial Intelligence at TUM.
Our third co-founder of Presize, Awais is from Pakistan and studied Biomedical Computing. As CPO, he is responsible for all product relevant topics like user experience.
3. The technology behind your start-up and your products is AI. How is this perceived by your stakeholders?
On the end-user and store side, the nature of the technology doesn’t play a particular role. At the end of the day, the end-user doesn’t care which technology is used to take their body measurements, the main thing is that it works.
The stores find the AI approach interesting and some already dealt with AI and Data Science before, but often their company-internal data doesn’t have the necessary quality to apply more AI.
The earlier you start investing in AI, the greater the sustainable competitive advantage in the future.
Where I particularly notice the influence of AI is in fundraising and recruiting. Of course, we are always looking for tech talents who also have job offers from large corporates. These talents often want to work with companies using cutting-edge technologies like AI.
AI also comes across very well with investors, as we use the data and technological know-how to build a sustainable competitive advantage.
4. How do you create value and help your customers with data/AI?
We create added value for three stakeholders with our technology:
- 💁 For the customer at home, we save time and effort, because the customer doesn’t have to go to the post office every time and return the goods if the clothes don’t fit.
- 👚 For online stores, we save money by reducing the number of returns. Each return costs about 20€ for the online stores, due to logistics and reprocessing of the returned goods.
- 🌎 We also save the planet, because less packaging is used and less transportation is needed. This reduces CO2 emissions.
5. Where do you see the opportunities and challenges of AI?
For me, AI is a wake-up call for digitization in many industries. Before you can even do AI or data science, the underlying data structure must be right. That means you have to link data silos with each other and guarantee the necessary data quality. There is an extremely large amount of potential in AI, but many other steps must first be taken.
A major challenge of AI is data protection in Europe. In my opinion, countries like the USA and China have a competitive advantage, because much more of this area is being researched and implemented.
We at Presize of course adhere to the GDPR and do not store videos and personal data. The face is pixelated during the recording and we only extract certain body measurements and match them to an artificially generated ID.
6. What invention do you hope to see in your lifetime?
I probably belong to the few Germans who see the car only as a means to an end. In general, I want to get from A to B as fast and efficiently as possible and therefore I’m very much looking forward to autonomous driving or even autonomous flying. I believe that this will bring a huge added value to many people including me.
7. If you could switch places with the CEO of a company for one day, which one would it be?
Even though I love my work at Presize, I would be interested in spending a day working as CEO of Wirecard. Especially now there’s a lot of need for clarification.