We transform design from the tools of your choice (Adobe Photoshop, Adobe WD, Figma, Invision)  into functional framework 

For designers

- Use your favorite tools (Adobe Photoshop, Adobe XD, Figma, Invision...) or handmade sketches
- Export your visual designs into Sketch files
- Convert simple design into ready to launch website
- Avoid hours on specifications
- Avoid iterations with developpers

For developers

- Convert designs into framework components
- Avoid obvious coding (regex, state variables, data population, form validation, basic wireframe, callbacks skeleton…)
- Automatically respect good practices
- Integrate classical plugins (SEO, interaction analysis…)

How does it work ?

Computer Vision

When you upload your design files, we recognize components thanks to our deep neural networks. We are able to identify them and to characterize them (size, colors, even movements).

Machine Learning

After component detection we are able to infer their roles and their interactions. "What should this button do ? Is it a subscription form ? Is it a plot component ?"

Website and app development

After understanding the design specifics, we generate automatically the corresponding code (vanilla HTML/CSS/JS, React Native, Angular and Vue are in our roadmap)

Frequently asked questions

We plan on supporting Angular, React Native and Vue, as well as simple HTML/CSS/JS solution.

We do not only generate “visual code” but we interpret what the design was made for and populate the source code of the app with already functional methods. Seriously, why are you still coding form validation ?

Sketch is a vector-based drawing tool with limited capabilities for code export. Viz2Code is not a design tool. It is a design scanner able to :

  • Tranform any visual into a layer based file (like Sketch file but not only)
  • Understand the objective of the component to generate relevant piece of code

If you provide us with an animated content (video or GIF) or a prototype which is available online, our tool is able to either decompose frame by frame the design and its moving parts or navigate through it. By using Viz2Code technology on each frame plus a smart tracking system we can generate basic animation code that you can edit.

The objective of Viz2Code is to accelerate the frontend development. If the design is complex, some work will be still needed from a developer point of view. In other cases, you must understand what is a website in its core to use our output files (HTML, CSS, Javascript or framework projects).

Still in Beta, Viz2Code will be launch in the beginning of 2020. You can sign up to be alerted when ready !

Do you want to work for us ? 😉 Drop us a line !

Our team

Anne Burrus


Anne holds a MIM from ESCP-Europe. Multi-time entrepreneur, ex-strategy consultant, she is passionate about artificial intelligence, web development and startups !

Joris Guerry


Joris holds an engineer degree from Arts et Métiers Paristech and a PhD from Ecole Polytechnique in Deep Learning for Computer Vision. He has been a web developer for 15 years.


For any questions or feedbacks, feel free to contact us or come meet us in Station F !

We are launching our beta version at the beginning of 2020 !

Drop us your mail and we will keep you updated when it is ready !