
AIGC MODE
Leveraging AIGC technology, reducing the threshold for visual creation through amazing and efficient AI content and tool capabilities, and creating a differentiated camera experience in Douyin (TikTok China).
Project Duration: April 2023 - October 2023
Results and Impact:
[Business Data]
-
Completed data verification for the MVP version. The follow - up rate of AI creation is 0.13%, higher than the overall average.
-
Launched the AI creation and AI portrait projects. During the event, the positive performance of submissions was remarkable, with an AI submission rate of 0.4%.
[Team Impact]
-
As the design owner, built a new design system and summarized design ideas in a short time, helping new team members integrate smoothly.
​
Strategy
Challenges
Facing brand
new technology, it was necessary to find the product direction and quickly launch and implement it within the Douyin platform.
Coordinating design team
responsible for onboarding new members and coordinating design work.
Design Insights
Capture the core features
What has the essence of AIGC changed? Solving productivity and generating novel effects in a short time → AI style special effects.
Priority
Given the complex technical implementation behind generative AI, where should we start to most quickly help the business launch and achieve results? → Scale impact within Douyin x factors liberating productivity x technical maturity.
Design Delivery
Explore opportunities for AIGC
powered creation and build a new design system.
-
Start from the submission genres and technical methods. Conduct modular sorting of the complex submission products and technical backgrounds to explore the opportunity points of AI - creation application scenarios.
-
As the design owner, coordinate the design priorities. Communicate closely with the business and R & D (RD) teams to launch and implement the AI - creation and AI - portrait projects.



Constructing a Universal AI Creation User Flow
-
Universal Framework: Emphasize discovery → use, providing universal structural support for the subsequent access of various AI - creation capabilities.
-
Simplified Operations: Different from most competitors that focus on "inputting prompts as the core experience", users can generate results more easily and quickly by selecting a style, taking a photo, and then choosing a prompt.
-
Inheriting the Follow - up Shooting Link and Amplifying Influence: Inherit the "special effect" follow - up shooting gameplay on Douyin. Users have access to create models from both the consumption side and the creation side.

AIGC Center

→Generate Avatar

→Edit details



Amplify the sense of surprise.
Support "Change" and "Random" functions.Discover more amazing effects.
Change Style
Radom






Summarize AI - creation design principles
Improve infrastructure, and help new members integrate quickly
-
Summarize the general design principles for AI creation. Key design points: highlight effects, input experience, surprise, and imitability.
-
Improve the infrastructure of image related functions: Be responsible for the first phase framework of text to image, image to image, AI image retouching, and digital avatars, helping new members quickly connect to the second phase projects.

Customized
AI Avatar

Edit:
Universal Framework

Consumption: Strengthen branding
