Galileo AI's Public Beta & Seed Round Led by Khosla Ventures
Today, we’re excited to launch the first public beta of Galileo AI, the prompt-to-UI generative AI tool for designers and builders.
Galileo AI emerged from a core belief: “AI will augment the creativity and productivity of designers, rather than replace them.” Our goal was to build capabilities beyond basic automation, and delve into the creative process that actually makes the user a better designer. We aim to enable the democratization of design, making it more accessible to a broader range of people, whether you’re a startup founder or software engineer. With Galileo, all it takes is an idea.
We’re also thrilled to announce our fundraising milestone: we have raised $4.4 million in seed funding led by Khosla Ventures, the first outside investor in OpenAI. It’s been wonderful working with Sandhya Venkatachalam from Kholsa. Our investors also include leading individuals in the design and productivity world. These include Julie Zhuo (former VP of Design at Meta), David Hoang (VP of Marketing and Design at Repl.it), and Howie Liu (CEO of AirTable), Dan Becker (Design leadership at Google), Anisha Jain (VP of Design at Cruise), Matt Robinson (CEO of Nested), Joel Hellermark (CEO of Sana), alongside many other esteemed industry leaders and experts.
A unique blend of AI + Design
Our journey started in late 2022, when we first came up with the idea for a text-to-UI tool. With Arnaud’s background as an AI researcher at Google and Faire, and Helen’s experience as a product designer at Facebook and Cruise, it’s the deep collaboration between AI and design disciplines that brought this product to life.
In the early stages, we quickly realized that out-of-the-box language models don't understand the subtleties of interface design. Therefore, we focused on training specialized models on tasks aligned with UI design standards. Helen's deep knowledge in design was able to drive significant improvements by training the AI on the concepts of components, colors, styles, and other design principles. Likewise, Arnaud leveraged his experience at Google where he learned to train state-of-the-art language models by creating custom datasets. This enabled the team to build a unique UI design dataset, which is key to the superior quality of UI generation.
In October 2023, we started rolling out a private beta, which quickly gained traction among product designers, startup founders, and software engineers. After an intense few months of iterating on their feedback, we’re ready to share the public beta with the world.
How it works
Think of Galileo as ChatGPT for interface design. The prompt can be "a fun, friendly app design for dog walkers" or “a dashboard UI for an e-commerce site” and you’ll be presented with multiple options for a visual layout in under a minute, with appropriate UI elements, images, and product copy.
One of the key learnings from the private beta was that users found a lot of value in generating multiple variants for the same design prompt. With multiple options, users can select their preferred version – and make further iterations and changes to achieve their final design.
As an example, one designer was developing an app to track cryptocurrency portfolios and they were able to produce all the key screens in less than an hour. They told us, “It would have taken me days to complete these screens without Galileo” – instead it took them less than an hour.
One of our R&D projects, Image-to-UI, revealed another great use case: generating designs from wireframes or screenshots. We've developed a feature capable of interpreting both text and images, enabling users to transform basic wireframes into high-fidelity designs. This was another highly requested feature from our beta users.
We're so excited to introduce Galileo 1.0 to you all. A huge thank you to our users, investors, advisors, and every supporter who helped us reach this point. Our seed funding allows us to work with AI visionaries, such as Khosla Ventures and our other new partners, to tackle greater AI challenges and stay at the forefront of AI and Design research.
Arnaud Benard & Helen Zhou