RingCentral is a leading provider of business cloud communications. It offers three key products  including RingCentral MVP™, a UCaaS platform including team messaging, video meetings, and cloud phone system. Video meetings include interactive online whiteboard for team collaboration.

The user, using a whiteboard, wants to replace his hand-drawn drawings with correct and beautiful figures to speed up work and increase the level of cleanliness.

Qualitative and quantitative research • IxD • VxD • Usability testing • Framer • Google Documents • Loom

Product Designer • Product Manager (Georgia) • Product Manager (USA) • Dev team • AI team • QA.

In order to tackle the aforementioned challenge, I conducted user research to gain valuable insights into the needs, preferences, and pain points of our users. The research primarily focused on understanding the current user interaction with the whiteboard, identifying their frustrations with hand-drawn sketches, and gathering their expectations for enhanced graphical elements.

— Gathered business and technical requirements (parameters for recognizing multiple drawings as a group).
‍
— Analyzed competitors in the market.
‍
— Conducted research on our demo model among company employees, dividing them into two groups: IT specialists and HR team.

— Accuracy of identifying drawings is around 80%

— Confirmed concern that presenting a large number of shape options from the edge of the screen slows down the workflow.

— When the whiteboard is crowded with content, users find it challenging to locate the recently replaced shape and continue their work. This issue was further investigated in the second phase of the research, conducted on a whiteboard with a large amount of data.

To explore interaction elements, the decision was made to conduct research in the realm of game design, specifically examining various menus found in action games where players need to make quick decisions. This approach aims to gather insights on effective design patterns for enhancing decision-making speed in user interactions.

After conducting thorough research and generating ideas, I created several interaction menu options and presented them to the team to gather their feedback and understand the technical considerations for each design. Through productive discussions, we narrowed down our choices to two promising options that we will further develop.

I took the initiative and created a dedicated icon library to replace hand-drawn sketches. The development team provided a list of drawings they used during the artificial intelligence training. To enhance the clarity and crispness of each shape, I applied a two-tone style.

During the design process, our AI team trained an artificial intelligence model, making it unnecessary to present 8 options. Therefore, I modified the interaction menu, still drawing inspiration from game design principles.

Developed a feedback collection feature from users to improve user-product interaction and continue training the AI using real-life examples, gradually expanding the library of recognized shapes.

Due to technical limitations of the project, we couldn't incorporate a feedback window after each meeting. Instead, users were prompted to provide feedback in specific situations:
— Multiple uses of the drawing reset feature.
— Inability to recognize a drawing.
— Anytime during their work

Users could express dissatisfaction with:
— Speed of operation.
— Accuracy of drawing recognition.
— Insufficient variety of shapes in the library.

Based on the initial feedback results, it was decided to add a set of shapes to the library, utilizing drawings provided by real users for training the recognition capabilities.

We rolled out the new feature to a focus group of customers who actively use the whiteboard. According to preliminary estimates, this feature expected that this feature will bring the company an income of 700 000$.