How Generative AI Transformed a Summer Event Design: A Step‑by‑Step Case Study

Preface

With the rapid rise of generative AI tools such as Wenxin Yiyan, ChatGPT, Midjourney, and Stable Diffusion, designers can quickly generate creative content by inputting prompts, images, and information, dramatically improving efficiency and diversifying designs for large‑scale activities.

Stage 1: Early Exploration

The summer activity aimed to satisfy students' travel and entertainment needs while using gamified operations to attract new users, increase activity, and boost DAU peaks. AI was employed in gameplay proposal, core IP setting, concept poster creation, and brand building to inspire concepts and accelerate implementation.

Gameplay Proposal

Using keywords like “summer,” “fun,” “parkour,” and “amusement park,” AI generated page mock‑ups that led to the selection of parkour as the main gameplay and “walk‑to‑earn” as a side mechanic.

Core IP Setting

Brainstorming produced dozens of animal ideas; AI‑generated images helped visualize options, and a corgi was chosen as the main IP after rapid visual comparison.

Concept Poster Setting

Five Chinese cities (Beijing, Xinjiang, Chongqing, Yunnan, Hainan) were selected as game backgrounds. AI generated city‑specific concept posters by combining regional keywords with visual composition.

Brand Building

AI generated slogans and graphic elements (stars, flowers) that were refined into the activity’s logo and auxiliary graphics, enriching the brand’s visual identity.

Stage 1 Summary

AI tools provided textual concepts, visual atmosphere, and rapid prototyping, enabling the team to establish a complete world view—character, setting, and activity—in a highly efficient manner.

Stage 2: Visual Implementation

After exploration, the activity entered visual implementation. Players control the IP to collect red packets across cities, unlocking new locations and a final “super‑large” reward.

Single‑Element Visuals

AI generated icons and city‑specific 2D elements (e.g., lanterns, dumplings) that were refined and integrated to keep asset size low while maintaining visual appeal.

Complete Scene Visuals

Complex visuals combined multiple AI‑generated images with activity elements for prize designs and resource‑slot entrances, creating cohesive and attractive scenes.

Reward Design

AI‑assisted postcard and wallpaper designs layered information, IP, and AI‑generated backgrounds, producing collectible prizes and themed phone wallpapers.

Resource Slot Entrance

AI‑generated splash screens and UI elements were blended with IP and reward graphics, ensuring a compelling entry point for users.

Stage 2 Summary

AI accelerated both single‑element and full‑scene creation, reducing design time while preserving quality and consistency across the activity.

Stage 3: Extension

AI‑generated assets were integrated into the MEUX component library, enabling team members to reuse them across various pages, further improving efficiency and visual consistency.

Conclusion

The project successfully delivered a multi‑city summer activity with over a hundred pages and assets, demonstrating how generative AI can boost design productivity while still requiring human creativity for strategic decisions and 3D realism.