TEAM AND ROLE
I worked as the UX Designer in a multidisciplinary team at HackHarvard 2022, collaborating with two developers and one business analyst. My responsibilities encompassed conducting desk research, user research, usability testing, and designing the complete user experience. Additionally, I played a key role in training our machine learning model, which was specifically engineered to accurately detect and analyze user emotions.
TIME FRAME
36 hours (Hackathon)
OVERVIEW
A video conferencing web application that provides real-time analytics on emotions expressed throughout meetings.
Empath.ly is an innovative video conferencing platform with integrated emotional analytics and real-time emotion identification. During video calls, our system analyzes participants' emotions sixty times per second through their camera feed. After each call, users can access a comprehensive dashboard displaying emotional data and actionable insights, including predominant emotions throughout the meeting, emotional patterns over time, and notable shifts or contrasts in participants' emotional states.
CONTEXT
Remote work has created an emotional disconnect in how we communicate with each other through screens.
The COVID-19 pandemic fundamentally transformed workplace communication, making remote interactions the new standard. However, this shift revealed a critical gap in human connection - our natural ability to read emotional cues and build relationships becomes severely limited through screens. Marketing teams and hiring managers now struggle to gauge reactions, assess engagement, and make informed decisions without the subtle emotional indicators that come naturally in face-to-face interactions. This challenge calls for innovative tools that can bridge the emotional disconnect in our increasingly digital workplace.
RESEARCH
Conducted desk research
Our business analyst and I researched to understand both the problem and the market better. We investigated why people prefer remote jobs and why certain video platforms succeed while others fail. This helped us identify opportunities for our solution.
Talked with users of popular video calling platforms
I spoke with people who regularly use Google Meet and Zoom to understand their experiences. Since these platforms are already widely used, we wanted to build something that wouldn't force users to change their habits. These conversations during the hackathon gave us valuable insights into user expectations.
Interviewed potential users across different roles
I interviewed people with various job titles from different companies to pinpoint exactly who would benefit most from our solution. These conversations helped us define our target audience more clearly.
INSIGHT #1
Our research revealed the need for balance between features and usability. When speaking with users of Google Meet and Zoom, we found they preferred Zoom's richer feature set but Google Meet's simpler flow. Users appreciated how Google Meet let them quickly join and use the platform for their intended purpose. In contrast, Zoom's interface felt cluttered with too many options, which reduced its ease of use.
INSIGHT #2
Through interviews with professionals across various roles and companies, we identified our primary target audience: HR professionals and marketing teams. These groups showed the strongest need for emotional analytics during remote interactions.
INSIGHT #3
Our breakthrough "Aha moment" occurred when we identified the core challenge: hiring managers and marketing teams lose critical emotional intelligence during video conferences. This emotional disconnect directly impacts their ability to build rapport with candidates, assess team engagement, and make strategic decisions based on authentic human responses - skills that are essential to their professional success.
PROBLEM STATEMENT
"How might we help marketing and HR professionals better understand and respond to the emotional dynamics of their virtual meetings?"
IDEATION & DESIGN
Designing for usability
My first priority was creating a user-friendly landing page. Our research showed that users struggle to accomplish simple tasks on platforms like Zoom because of cluttered interfaces. This wastes valuable time for HR and marketing professionals who need efficiency. I designed our landing page with clear pathways to start or join meetings without distractions.
Leveraging Familiar Design Patterns
Since users are already accustomed to platforms like Google Meet and Zoom, I recognized the importance of working with these established mental models. Rather than forcing users to learn an entirely new system, I designed our video calling interface with familiar elements inspired by Google Meet. This approach reduces the learning curve and helps users feel comfortable immediately.
Biggest Challenge: Effective Data Visualization
The most significant design challenge was presenting emotional data in a meaningful and accessible way. We needed visualizations that anyone could understand at a glance, regardless of their data literacy. My team and I collaborated at the whiteboard, exploring various data representation methods. We examined different visualization techniques to find the perfect balance between comprehensive insights and clarity for our users.
Beyond Data: Understanding User Needs
While exploring different ways to present emotional data, we realized we needed to dig deeper. Simply showing data wasn't enough—we needed to understand what specific insights would actually help our target audience make better decisions. To answer this crucial question, I conducted additional targeted research with HR and marketing professionals to identify the emotional patterns and indicators that would be most valuable to them in their work.
Respecting Users' Time Constraints
During follow-up research, I discovered that many users have back-to-back meetings. This means they don't always have time to review detailed data breakdowns. Therefore, I designed a pie chart dashboard that appears immediately after meetings end. This provides users with a quick understanding of the emotions expressed throughout the meeting, allowing them to quickly grasp key insights and transition to their next meeting.
Translating Research Insights into Dashboard Features
I designed the dashboard with simplicity and scannability as core principles, ensuring users can easily find the data they need. The dashboard presents information through three distinct visualization methods to address different user needs:
Quick Overview: A pie chart provides an immediate emotional summary for users with limited time between meetings.
Detailed Timeline: A graph displays the flow of emotions throughout the meeting, helping users understand emotional patterns over time.
Contextual Highlights: The application transcribes meetings and highlights specific moments where certain emotions peaked, providing actionable insights tied to actual conversation points.
