Our patent-pending technology

Get results that would have required a 50k€ hardware scanner just one year ago. We generate a 3D model from a smartphone video, providing much greater insight into the human body than just two photos or a few questions.

Tech-driven value add

High accuracy

As accurate as a professional tailor who would need 10 minutes to take manual measurements

High robustness

Our video approach produces consistent results for the same body

Largest dataset

We combine and augment multi-modal data from several sources and acquire new labelled data every week to retrain our algorithms

Constantly learning

Past returns from similarly shaped users help us improve size future recommendations

Based on groundbreaking research

Our team members have academic backgrounds from some of the best universities in the world (Stanford, Cambridge, TUM). We developed our unique technology based on state of the art research, and collaborate with high-profile partners from academia.

100% browser-based

Not just another app you download once and never use again!

We are completely browser-based, which means our technology works across all devices (smartphones, tablets and desktop) and doesn’t require users to download an app. We believe our browser-based web app is superior to mobile apps, since downloads kill conversion: People don‘t want another app just to scan themselves once.

Personal 3D Avatars

Until recently it was unthinkable to reconstruct a realistic 3D model based on just a simple RGB smartphone video! Before presize, to get a 3D model with comparable quality required expensive hardware with multiple cameras and sensors. We are currently using these models to extract user body measurements and starting in 2020, models will be displayed to end users – this is a powerful visualization since we capture photorealistic detail and texture, increasing "perceived accuracy". We can also use these models to deform clothing (that's why users don't need to undress).


presize ID

The personal key for the end-user’s perfect fit. Once body measurements are extracted, we delete the associated video and never collect or store names. A presize ID is anonymous, shareable and does not require user registration – we simply store the data up to 60 days within cookies. Alternatively, the user can register to keep their ID forever. Your recurring customer base will continue to expand as users are guaranteed the perfect fit every time (lock-in effect).

Personalize the user journey

Users get frustrated when they go to select their size for a product they like, only to find it unavailable. With presize, you can choose to display only the products where the shopper’s size is available and even order products by body shape/type. We open you to unprecedented, hyper-focused retargeting options based on user bodies. Need to shift leftover stock in unpopular sizes?


You can easily integrate our service via Google Tag Manager on any website or opt-in for custom integration using our API. We provide all the code snippets and widgets, you just paste and customize!

Core Services

Measurement Extraction

Smartphone video in, body measurements out, completely anonymized.

Size Recommendation

We match the user’s body measurements to product specific data and existing user feedback to recommend the best available garment size.

Product Sizing Data

We standardize, digitize and structure any product data provided by the shop and combine multiple data sources.

Brand-level sizing charts

e.g. the Adidas sizing chart for male upper bodies

Product-level sizing charts

e.g. the data for a specific pair of Adidas Originals Sweatpants

Existing database

We crawl and scrape existing sizing data

Poor product sizing data?

Imperfect product data can be improved and supplemented by presize.

Learn from past returns

Return data is used to fine-tune our size recommendation algorithm

Digitize clothing

You can provide us with physical products for our team to digitize the sizing data

Learn from user behavior

User sizing decisions diverging from our recommendations may suggest poor product data which produce radically different sizes typically purchased by the user

Ready to measure with presize?

Try Now!