We had the distinct pleasure to interview award winning designer Ruoyong Hong about their original and creative Personal Email Assistant App design work. Ruoyong Hong is an award-winning communication designer, that makes our world a better place with their original personal email assistant app works.
Ruoyong Hong ("RH") answered Design Interview ("DI") questions about their award-winning work Wel AI Personal Email Assistant App. Interview questions prepared by Frank Scott.
RH : The inspiration behind the project “Wel AI” comes from observing that people often struggle with writing proper emails or messages for various situations. There are many potential solutions to this problem, ranging from searching for templates to purchasing writing services. However, one idea stood out – an AI assistant for composing emails. While this is not a new concept, existing AI software still falls short in providing substantial help. Simply detecting tone and replacing words is not enough. We began with thorough user research and open conversations with various individuals. Through this, we learned about the current state of writing assistant apps and services, and we identified significant opportunities in this market.
RH : The goal was to create a mobile-based writing assistant that integrates AI into daily email processes, facilitating analytical reading and offering a conversational approach to seamless email composition, guidance, and drafting. The design stands out for fostering a flexible and respectful collaboration between humans and AI, providing assistance that feels just right. It cultivates a personal writing style and a social relationship library, benefiting both the user and the AI, encouraging growth and improvement like a supportive buddy in life.
RH : I plan to continue researching and exploring how the latest AI technology can be integrated into this design concept. Building a personal contact library and offering writing assistance that is sensitive to people with different national backgrounds are key areas of focus. I also aim to promote “Wel AI” to a diverse audience to identify customization opportunities and ensure the app meets the needs of users from various backgrounds.
RH : The design was initially conceived as a UX research initiative focusing on trust in human-AI collaboration in the summer of 2018. It has evolved significantly since then. With insights from the research phase and the expertise of AI scientists, I spearheaded the development of this proof-of-concept app as the lead designer in 2022. Therefore, it took approximately 3-4 years from concept to design, predating the popularity of ChatGPT.
RH : At the time of the initial concept, suggestion-based AI systems were insufficient, and people continuously sought further help from AI for email writing. There was a growing desire for AI to assist in a customized way. However, when we turned this idea into a product, the issue of trust in AI emerged. Would people want a "robot" to represent them? How far would they like AI to help, and in what ways? This led me to explore a design that could bridge this gap and establish a new pattern of interaction with AI in writing.
RH : This design distinguishes itself from two main concepts. Firstly, traditional email writing assistance products focus primarily on word or emotional detection, with the best ones helping to rephrase individual emails. In contrast, our design leverages previous conversations and emails to build a personal library, enabling users to write in their own tone and cater to their social connections. Secondly, unlike AI conversation products, Wel AI introduces a more integrated interaction by providing AI assistance step by step, rather than controlling everything through typing or talking to the AI. Users can choose the level of assistance they feel is just right from Wel AI.
RH : I utilized a variety of design tools to aid in the development and execution of the design. Using Figma and Photoshop, I created an intuitive interaction flow and polished visual system. The project followed a human-centered approach, starting with user research, transitioning to design exploration, and concluding with prototype testing. The product integrates GenAI for text tone and intent analysis, drawing insights from prompts in a writing and relationship library. It reviews users' daily actions and selections to continually optimize performance and closely align with their personalities. Key design tools included user research methods such as interview guides, note-taking apps, and audio/video recording devices.
RH : Beyond its AI features, the most unique aspect of Wel AI is the intuitive swipe interaction logic and clear visuals that separate user input from AI input. Wel AI uses tags throughout the app to make text-heavy tasks easier. On launch, it shows a brief of emails with tags indicating their purpose. Swiping up reveals a detailed list, while swiping left or right transitions between Assistant and Librarian modes, representing work and training mindsets. In the email view, swiping up displays in-depth analysis or seeks AI guidance. The use of blue and white shades subtly communicates the depth of AI involvement, blending trust with clarity, much like a supportive buddy in life.
RH : Yes, my design is heavily influenced by data and analytical research. We conducted user interviews using AI prototypes alongside scientists to understand user needs and how users would interact and grow with AI, as well as to assess the feasibility of the technology. We invited more than 10 users from diverse backgrounds to try and use the prototype of this design. Additionally, we conducted multiple rounds of quantitative research to gather detailed design feedback. Our research included a diverse group of email users, ranging from students to professionals. We utilized tools such as user personas, flow maps, affinity maps, and process mining to reveal how fluctuating AI assistance and trust are based on user confidence and task context familiarity. Combined with a human-centered UX design approach, these insights significantly shaped the core concept of our design.
