02 REMESH
PRODUCT / AI / SAAS
COMMON TOPICS
Disclaimer
The views expressed in this case study are my own and do not reflect the views of Remesh, Inc. All design assets and concepts presented are the property of Remesh, Inc. This case study is intended solely to showcase my design process.
About Remesh
Remesh is a SaaS tool designed to help users instantly understand a population by leveraging AI technology. The platform enables moderators to host live conversations at scale and analyze responses in real time. To see the platform in action, watch the video here.
My Role
As a senior product designer, I was responsible for Designing the live conversation experience for open-ended questions, enabling researchers to instantly process and understand 1,000 responses in real time.
The User Problem
Researchers needed to read and analyze up to 1,000 responses within a two-minute window as participants submitted answers during a live conversation. This overwhelming volume of qualitative data made it difficult for moderators to manage the session and engage effectively in real time. As a result, key moments to ask follow-up questions or gain deeper insights were often missed.
Our user research revealed that Jenny, a typical moderator, constantly sought to identify common themes and topics emerging from conversations, while simultaneously managing live responses. The process was not only time-consuming but nearly impossible without missing critical insights.
Hypothesis
By surfacing the most frequently mentioned topics during live conversations, we could significantly reduce the time researchers spent analyzing responses, enabling them to focus on moderating and engaging with participants in real time.
Ideation
I led a collaborative whiteboarding session with engineers, product managers, and other stakeholders to brainstorm ideas and develop prototypes for a solution. Together, we explored ways to streamline the process for identifying common topics in real time.
The MVP Solution
We developed the Common Topics feature, designed to help researchers quickly identify and understand the emerging topics from live conversations. Powered by NLP, the system extracts key phrases and groups responses with similar topics, allowing researchers to focus on key insights more efficiently.
Real-time Topic Highlights
As responses flow in during the live conversation, the All tab displays a stream of responses, similar to a news feed. At the same time, the algorithm highlights frequently mentioned topics.
Common Topics Snapshot
The Common Topics Snapshot provides an overview of the most mentioned topics. It displays one response with the highest agreement score for each common topic, helping researchers quickly gauge the direction of the conversation.
• A ranked list of common topics appears on the left, organized by frequency.
• Clicking on any topic reveals all related responses, with key nouns and adjectives underlined for easier reading.
Improving the Algorithm
Users can remove topics that are inaccurate or insignificant. This helps them better organize the topic list and also aids the algorithm in learning and improving over time. By incorporating user feedback, the system becomes more refined and accurate with each session.
Outcome
Time Efficiency: Researchers now spend significantly less time analyzing large volumes of data in real time.
Improved Engagement: Moderators are more confident in managing live sessions, as they can quickly understand and react to audience responses.
Continuous Improvement: User feedback is directly integrated into improving the NLP algorithm, ensuring the tool becomes more accurate with each session.