What is AI XiaoKuai?

I'm an AI virtual user on Kuaishou. Powered by Kuaishou's self-developed large model, I can interact with everyone in the comments section of posts! 🤩

Who are your primary users?

The majority are young users under the age of 23! 👱‍♀️

Compared to the general commenting user base, the DAU of AI XiaoKuai skew significantly younger, within the 12-23 age bracket.

Users aged 12-17 are the core driver of this platform, contributing nearly half of both its user base and total interaction volume

DAU in the broader comment

DAU who @ AI XiaoKuai in comments​

What do people generally ask you to do?

Everyone usually chats with me in the comments section and asks me questions 🤨

They also ask me to help generate some images 😊

User interactions with AI XiaoKuai are primarily text-based,

with the largest share being knowledge-based Q&A, everyday life advice, and celebrity gossip.

User interaction modalities with AI XiaoKuai

Text-based interaction modalities with AI XiaoKuai

Why embed AI Xiaokuai into the social message module?

Simply because there's user need. Even before I could support private messages, many users sent them, despite getting no response.

Plus, it holds value for our business metrics.

The group aged below 18- are the most active users of AI XiaoKuai.

They are the most frequent users of private messaging and are particularly engaged with interactive AI functions.

How was you integrated into the messaging module?

I greet first-time users on the session page with hints about my capabilities.

Recommended prompts help users overcome the blank canvas syndrome, while also showcasing the AI's capabilities and driving the conversation forward.

These suggestions are typically displayed as a list of 3 to 5 options.​​

I also provide ready-to-use command templates for common user tasks. Users can simply follow the steps and input their own content to get tailored results! 🎯

AI pre-filled structured templates simplify the creation of precise and professional instructions

After obtaining results, users can further expand their understanding by exploring related questions.

The second-order association lets users refine existing drafts instead of starting over, streamlining the process and avoiding parameter overload.

Leveraging typical social media user habits, any generated result can be applied instantly with a single click for maximum convenience.

Depending on the type of generated outcome, in addition to basic universal operations, corresponding quick actions are also provided.

Text>Copy

Image>Share

Avatar>Set

Background Image>Set

How did you go in the social messages? 😊

Hahaha, achieved significant data-driven gains!​​

Data benifits

Next-day retention rate comparison

更多项目

Create a free website with Framer, the website builder loved by startups, designers and agencies.