How Older Adults Use Online Videos for Learning (CHI '23)

Mobile Interface Design for Low-literacy Populations (SIGINT '12)

Older Adult’s Voice Assistant Use for Health Information Seeking (TiiS, '22)

Physically Disabled Players in Movement-Based Games (CHI '22)

Pay Attention, Please (CHI '19)

Keep it Simple (CHI '20)

User Expectations of Conversational Chatbots (DIS '21)

Transform any paper into

design cards.

Harnessing generative AI models, our system automatically transforms a PDF paper into design cards.
You can further customize these cards simply by providing your own design context.

Upload a PDF paper

Upload an HCI paper to the system

Select a paragraph of the paper

Then, our system automatically segments each part of the paper with the heading. From the list of paragraphs, select the paragraph(s) that you want to generate design cards from. Then, the system automatically generates a double-sided design card, with eye-appealing visuals included

Generate design cards

Provide your own designing context (optional)

If you want to position the design cards into your own design work, simply describe that in the text field, which will result in generating another customized design card side that is situated in your own design context

Generate customized design cards

Who made this?

From Paper to Card: Transforming Design Implications with Generative AI

Donghoon Shin, Lucy Lu Wang, Gary Hsieh

CHI 2024 (to appear) PDF Project website


  • While design cards are accessible and provide more prescriptive insights to practitioners, one challenge with relying on using design cards to communicate academic findings is how these cards should be generated. It is unrealistic to expect researchers to do it, as HCI researchers describe the lack of time and competing responsibilities that prevent them from doing more outreach work. Thus, we built this interface to scale up the creation of design cards using generative AI.

  • Design cards are tools that facilitate the design process, and can contain several types of design knowledge—from human insights to material & domain knowledge. There is a long tradition of creating design cards to communicate research insights.

  • There is no specific requirement regarding the paper topic / discipline for generating cards; however, we have tested and demonstrated empirical efficacy of our system using HCI papers.

  • The generated cards and the provided feedback will be saved in our system, and they may be used for the future analysis.

  • Yes! The analysis of our system will be published in CHI 2024, and the codebase can be found here.

This work is supported by an from Microsoft