Play with generative AI models to bring your imagination to life.

In early 2021, the team at WOMBO released the original WOMBO app which allowed users to upload a selfie and turn it into a dancing video using AI. The app gained virality quickly, and the WOMBO team decided to later expand their hold on AI creation tools by later releasing their second hit, Dream, in November 2021.

I was fortunate to join the team as one of their founding designers around July 2021, working on marketing, animations, and quickly transitioned into the product design team.

Team

Daybreak Studio, Peijia Ding

Role

Product design (iOS/Android), visual design, animations

Social experiences

Showcasing the community

The discover feed showcases artworks created by the community. We kept in mind a modular feed, broken up into sections where we can highlight different types of features in Dream. The simple layout and UI is timeless, allowing us to easily adapt as we expand more community-focused features.

The creation tool

Dedicated to creation

Being early believers in the AI space, our small design and engineering team have managed to both set and exceed the standard for AI artwork tools. Some of my favourite accomplishments include improving and launching new features to our core creation flow.

Create with images

Dream's interpretation of input images was one of the first of its kind in the AI art world. As it was a completely new flow for our users to learn, we designed the experience to be clean, simple, yet still informative.

The influence slider

One main challenge was finding a way to communicate the image influence. After many iterations, we settled on a three-point slider that communicates the influence both visually through blurring the preview photo, and in easy-to-understand language.

The impact

The release of our new input images features released in early March of 2022, and took TikTok by storm. Dream quickly saw success and climbed the charts to #1 on the App store in over 93 countries.

Expanding our toolset

Before: The classic linear flow

Previously, the creation flow only contained of a user input, generation, then the results page to display the user's artwork. In fall of 2022, we decided to integrate more tools that allow for our users to gain more control of the artworks they generate.

After: Introducing editing & focused states

User testing and customer feedback led the product team to explore more editing features for dream. With more features coming into the picture, we also put an emphasis on focused states in order to not overwhelm the user.

A focus on prompts

We decided to expand our prompt box when activated to allow users to easily edit and create longer prompts. The visual weight of the expanding box also creates a focus state, so users are not distracted by other parts of creation when writing their prompt. The introduction of enhancing prompts was launched with this update, which uses AI to give users a more descriptive version of their inputted text.

Create more in less time

Multiple outputs allows users to see four different interpretations of their prompt and then finalize one to post onto Dream.

The release of multiple outputs quickly became the top feature to convert free users to premium users.

Reflection

Being in a small and close team allowed me to have a voice in every part of the product– from overall product direction and ideation, to maintaining strong communications with our development team.