Jombay- AI-Powered Engagement Solution
India’s first AI powered Engagement Survey
Role
UX UI Designer
Team
Product Manager, Lead Designer, Engineers, CPO
Duration
Feb - Sept 2024
Portfolio by AARTI SAPKAL
UX UI Designer
User Flow to record response
Due to NDA restrictions, all the details about the project, design and design process cannot be shared. Here are some glimpse of the design.
Recording response
Simple user interface
user can record voice response here.
can delete it if something went wrong
Stop when done recording.
1.
Listen to your recording
User can replay the recording to ensure what they are submitting.
User can either delete the response or submit response.
2.
Transcript loading...
Once user submits the recording, system will generate a transcript.
By the time system generates the transcript user can move to next question and edit the transcript later on.
3.
Review Transcript
User can edit the transcript once ready
4.
Break..
It could be overwhelming for the user to answer 3 subjective questions in one go
So we introduces small breaks
User can save progress, take a coffee break and come back to complete remaining survey.
5.
Testing
Voice-to-text feature Prototype Testing
I identified 2 major issues in this prototype:
How I solved it?
Voice-to-text can make feedback experiences more human.
Collaborating closely with PMs and engineers helped balance feasibility with design quality.
Even small design tweaks (like editable transcripts) can build user trust.
Privacy Concerns:
Users assumed their voice recordings were being submitted instead of the text transcript, leading to fears that their managers could identify them.
Lack of Clarity About Re-recording:
Several users were unsure whether they could re-record their responses or edit the transcript before submission.
I introduced a clarity pop-up that appears the first time a user clicks Record Response.
Can record multiple times and edit the transcript, and
Only the transcribed text (not the audio) is submitted, ensuring full anonymity and control.
Dashboard and AI action plan
To make sense of all the responses, I designed an analytics dashboard and Ai action plan
Dashboard displayed AI-driven thematic analysis (Positive, Negative, Neutral) along with engagement score and other key metrics
AI action plan suggested focus areas and solutions based on survey results
Managers could customise issues, edit AI suggestions, and track progress
Workplace of Winners (WoW) Reviews
Organisations can publish Positive
employee stories on WoW page as
authentic testimonials, enhancing
employer branding.
Impact
The solution became the foundation of Jombay's "Workplace of Winners" program
One of India’s first engagement surveys with AI powered action planning.
Adopted by 300+ organisations in its first year
#1
300+
What I Learned 💭
Overview
About the Project
As part of Jombay's mission to build "Workplaces of Winners", I designed India's first AI-powered engagement survey, an integrated solution that helps organisations understand employee sentiment through richer, more emotional insights.
The Engagement Solution combined three parts:
Problem
How might we encourage employees to share their real stories and emotions easily?
Open-ended questions in employee surveys often go unanswered or receive brief, generic responses.
Research
Understanding Employee Feedback Barriers
I studied existing employee-survey tools, did some secondary research and also conducted short interviews within our organisation:
Ideation
Exploring Multiple Approaches
I brainstormed multiple ideas to encourage storytelling among employees:
Prototype
Voice-to-text feature to record responses
I designed a clean, distraction-free interface where employees could
Record their responses verbally,
Edit the auto-generated transcript before submitting,
Re-record multiple times until satisfied
While exploring these options with team, we realised that most features could influence or bias the employee’s response.
To maintain authenticity and capture true emotional nuance, we chose to prototype the voice-to-text feature enabling users to express freely in their own words and tone.
Users find the the open-ended questions intimidating
Users consider this as just another task to finish and they provide generic answers
No way to capture voice of employees literally or emotionally
Key insight: People are more open and expressive when speaking rather than writing
Engagement Survey
To capture employee stories through voice
Analytics Dashboard
To visualise organisation-wide insights with thematic analysis
AI-Powered Action Plan
To turn data into meaningful action to improve the employee engagement
What You'll See in This Case Study
Problem
1
Research
2
Ideation
3
Testing
4
Impact
5
Story starters: short, conversational phrases like “I feel most valued when…” or “One thing I wish was different…” to lower hesitation
Mood picker — users could select an emoji or emotion tag that best represents how they feel
Voice-to-text input for more natural and effortless sharing
Emotion sliders to capture tone and sentiment
