UX / Product Designer · Noida, India

Rohit Kumarhelps people understand
their data and act on it.

I am a UI/UX designer with 2.5 years at EY GDS, where I have mostly worked on enterprise AI tools and dashboards that help teams make faster decisions. I care a lot about reducing clutter and making sure the most important thing on a screen is actually the most visible thing.

AI/GenAI UXDashboard designEnterprise workflowsDecision-support UIDesign systemsData visualisation
2.5+
Years at EY GDS
8+
Products worked on
GenAI
Primary domain
PrepMind
Personal project
UX Case Studies02 personal projects
CS 01 · Personal project
Case study
PrepMind — MBA Prep Platform
UX ResearchIAWellbeing UXFigma
Wellbeing-focused MBA exam prep platform I designed from scratch. Restructuring the IA cut navigation complexity by around 30% and noticeably reduced the feeling of overwhelm during study sessions.
CS 02 · Concept project
Case study
¢
↓18%
₹12.4k
spent
82%
on track
₹3.6k
left
Swiggy · Food
₹340
Uber · Travel
₹185
Cent — Financial Management App
AI UXFinTechIndiaConcept
An AI-powered expense tracker built for how India actually pays. Real-time UPI and SMS detection, 1-tap categorisation from the notification itself, and an AI that learns your habits so eventually you do not have to do anything at all.
Office Projects05 projects · EY GDS
These are projects I worked on at EY GDS between 2023 and now. Most are internal or client-facing enterprise tools so I cannot share screens publicly, but I am happy to walk through any of them over a call.
01 / 05 · EY GDS
★ Highlight
GenAI Product Suite
AI/UXEnterprise5 tools
Five AI tools built for enterprise consulting workflows. The tricky part was designing around AI-generated outputs where the user needed to trust what they were seeing.
02 / 05 · EY GDS
AI Tool
High risk
High risk
Medium
Low
Health Claims STP Agent
AI ToolHealthcareRisk UX
Insurance claims reviewers deal with massive volume every day. I made sure the high-risk claims are impossible to miss and the next action is obvious without needing to read everything.
03 / 05 · EY GDS
Dashboard
+18%
Revenue
92%
Retention
3.2x
ROI
Sales Dashboard — CPG
AnalyticsCohort UXKPI design
A CPG sales dashboard for field reps and strategy teams. These two groups look at the same data very differently, so figuring out the right hierarchy for both took a lot of iteration.
04 / 05 · EY GDS
Engagement
File access
Workflow
Data clarity
Benchmarking Tool
EnterpriseData UXWorkflow
Data was technically accessible but practically hard to use. I rebuilt the IA and added access-controlled uploads to get users from seeing data to acting on it with far fewer steps.
05 / 05 · EY GDS
Margin & Portfolio Analysis Tool
Finance UXKPI dashboardsMulti-BU
Cross-business-unit margin tracking at CFO level. When you have multiple business units and overhead allocations to compare, tables alone do not cut it.
Unit A 35%
Unit B 25%
Overhead 15%
Other 25%
About Rohit

I am a UI/UX designer at EY GDS in Noida, where I have spent the last 2.5 years working on enterprise products — mostly AI tools, dashboards, and internal platforms used by consulting teams. It is the kind of work where the data is complex and the users are busy, which I genuinely enjoy figuring out.

My degree is actually in mechanical engineering from NIT Kurukshetra, which sounds unrelated but it shaped how I think. Engineering taught me to understand systems before touching solutions, and that is still how I approach design problems today.

I am a detail person. I notice when padding is off by 2px.
I prototype until it feels right, not just until it looks right.
I always ask what the user is supposed to do next.
I would rather simplify a flow than add a tooltip explaining it.

Outside of work, I travel whenever I can, I have been watching anime for years (used to make edits too, which is how I got into video editing), and I spend time in After Effects and Premiere Pro just for fun.

Hobbies
✈ Travel
New places genuinely change how I see design problems.
○ Anime
Been watching for years. Anime edits got me into After Effects first.
▶ Video editing
After Effects and Premiere Pro, mostly for personal projects.
⬢ Prototyping
I have a hard time calling something done until it is fully interactive.
Skills
Dashboard and analytics UX
AI and GenAI product design
UX strategy and IA
Wireframing and prototyping
Interaction design
Design systems
Usability testing
Data visualisation
Motion and video editing
Tools
FigmaFigJamAdobe XDPhotoshopAfter EffectsPremiere ProMiroNotion
Education
B.Tech — Mechanical Engineering
NIT Kurukshetra · 8.16 CGPA · 2019–2023
Class 12th — CBSE
Bharti Public School, Narnaul · 86% · 2019
Class 10th — CBSE
Bharti Public School, Narnaul · 9.8 CGPA · 2017
Certifications
Figma UI UX Design Advanced
Daniel Walter Scott, Udemy · Apr 2025
Design Thinking with ChatGPT
SeaportAi, Udemy · Feb 2025
Design Principles: an Introduction
UC San Diego, Coursera · Feb 2025
UX Design Fundamentals
Joe Natoli, Udemy · Sep 2024
Figma UI UX Design Essentials
Daniel Walter Scott, Udemy · Mar 2024
Contact

Always up for
good design conversations.

I am based in Noida and currently open to new opportunities. Whether it is a role, a collaboration, or just a conversation about design, feel free to reach out. I try to get back within a day.

← back to work
Case Study 01 · Personal project

PrepMind — MBA Prep Platform

Role
UI/UX Designer
Type
UX Research, IA, Product Thinking
Tools
Figma, ChatGPT, Google Forms
Project
Personal project
Overview

PrepMind is a personal project I have been working on. An MBA exam prep platform designed around the idea that studying is stressful enough without the app adding to it. Most prep platforms throw everything at you at once. I wanted to design something that felt more like a structured plan than a content dump.

Key design decisions
Impact
30%
Less navigation
complexity
40%
Less decision
fatigue reported
4 modules
One connected
platform
Live prototype
View prototype in Figma ↗
High fidelity web app UI
UX thinking process
01
Introduction
What this project is and why it exists

PrepMind is a mental-wellbeing-first MBA exam preparation platform. Most prep platforms give you content and leave you to figure out the rest. I wanted to design something that actively reduces the stress of preparation, not just the academic difficulty.

The goal was simple: help aspirants prepare consistently without burning out. The differentiator is the wellbeing layer — 1-on-1 mentor access and therapy sessions built directly into the product. No other coaching platform does this.

What makes PrepMind different
Every other CAT platform focuses entirely on content volume. PrepMind is the first to treat mental wellbeing as a core product feature, with 1-on-1 mentoring and therapy sessions built into the main experience.
02
Problem
What MBA aspirants actually go through

The problem is not that students do not have enough resources. It is that they have too many, with no structure to use them, and no support when things get hard mentally.

📚
Resource overwhelm
Too many books, courses, YouTube channels. No clarity on what to do first or what to skip.
📅
No structured plan
Most students start strong but lose direction within weeks because there is no daily guide.
🤯
Decision fatigue
Deciding what to study, when, and for how long every single day is exhausting.
🔥
Anxiety and burnout
Emotional difficulty during prep is just as real as academic difficulty. Platforms ignore this entirely.

Design challenge: How might we create a structured, low-stress preparation system that improves consistency without overwhelming users?

03
Research
Surveys and behavioural analysis

I ran a survey targeting CAT and MBA aspirants at different stages of preparation. The goal was to understand not just what they needed academically, but what made them quit or slow down.

📉
Motivation drop
Users feel lost after the initial excitement wears off, usually within the first few weeks.
🔄
Inconsistency
The absence of a daily structure is the biggest driver of inconsistency, not lack of intent.
Feature overload backfires
More features and tools increased anxiety. Users want guidance, not options.
💬
Emotional burnout
Students rated emotional difficulty on par with academic difficulty, but no platform addresses it.
04
User persona
Who I was designing for

Based on research, a primary persona emerged: non-engineers with strong verbal skills, struggling with quant confidence and study consistency.

Background
Rohit, 23
Non-engineer
Preparing for CAT 2026
Working and studying
Goals
Score 95+ percentile
Stay consistent
Avoid burnout
Build quant confidence
Pain points
Quant anxiety
Overthinking plans
Hard to stick to schedule
Too many choices

Rohit is strong in verbal but loses confidence the moment quant appears. He is not lazy, he overthinks and gets paralysed by choice. He needs the app to tell him exactly what to do today, not give him another menu to navigate.

05
User journey
How a typical aspirant moves through preparation

Mapping the journey revealed that the problem is not at the start. Aspirants begin motivated. The real friction points are in the middle stages, where structure breaks down and emotional difficulty sets in.

Stage 1
Motivation
High energy. Enrols in courses, buys books.
Stage 2
Overload
Too much content. Unsure what to prioritise.
Stage 3
Confusion
Makes and breaks study plans repeatedly.
Stage 4
Inconsistency
Miss days, guilt builds, momentum drops.
Stage 5
Burnout
Either quits or pushes through alone.
Opportunity
Intervention
Introduce structure and reduce decisions at every stage.
06
How might we
Turning research into design opportunities

Each question came directly from a pattern in the research data, not invented in a workshop.

  • How might we reduce decision fatigue in daily study planning so students do not have to think about what to do next?
  • How might we create a system that guides users through their day without requiring them to build discipline from scratch?
  • How might we integrate mental wellbeing into the preparation experience so it feels like support, not a separate product?
  • How might we simplify navigation across multiple modules so the app feels like one coherent experience?
  • How might we make quant feel less intimidating for students who have already convinced themselves they are bad at it?
07
Feature ideation
From insights to product decisions

Every feature was tied back to a specific research insight. Nothing was added because it felt like a good idea, only because a user behaviour or pain point demanded it.

Core
Diagnostic test
Starting point for every user. Builds a personalised baseline before recommending anything.
Core
Smart study planner
Auto-generated daily plan based on diagnostics and exam date. No manual planning required.
Core
Daily task dashboard
Tasks are pre-decided. User opens app and starts, no decisions needed.
Core
Mock tests and analysis
Full-length mocks with sectional breakdowns and weak area identification.
Differentiator
1-on-1 mentor sessions
Live sessions with mentors who have cleared CAT. Strategy, motivation, quant doubt-solving, all in one place.
Differentiator
Therapy and wellbeing
Access to counsellors during high-stress periods. No other prep platform offers this.
08
Information architecture
How the platform is structured

The IA was designed to feel like four clear rooms, not a drawer of tabs. Each module has a single job and does not bleed into another.

Dashboard
Daily plan
Task list
Streak tracker
Mood check-in
AI nudges
📅
Planner
Study roadmap
Milestones
Adaptive replanning
Weak area schedule
Exam countdown
📝
Test management
Mock tests
Sectional practice
Post-test analysis
Accuracy trends
Time data
📈
Analytics
Performance tracking
Benchmarking
Weak topic heatmap
Time-on-task
Improvement rate
Wellbeing — the fifth module
Wellbeing sits alongside these four as a primary navigation item. Students access 1-on-1 mentor sessions and licensed counsellors the same way they access a mock test. This is what no other platform does.
Personal project, not affiliated with EY. Full case study and Figma file available on request.
← back to work
Case Study 02 · Concept project

Cent — Financial Management App

Role
UI/UX Designer
Type
AI UX, FinTech, Concept
Tools
Figma
Project
Concept project
Overview

Cent started from a frustration I think a lot of people have. You spend money on UPI all day, get SMS alerts you immediately forget, and then feel vaguely guilty at the end of the month without knowing where it all went. I wanted to design something that tracked your spending without you having to think about it. The whole project was built around one idea: what if financial awareness was just something that happened, rather than something you had to do?

Key design decisions
Impact
85%
Less manual
data entry
3x
Daily interaction
vs other apps
<30%
Drop-off vs 70%
industry average
Live prototype
View prototype in Figma ↗
Financial management app UI
UX thinking process
01
Introduction
What Cent is and why I built it

Cent started from a frustration I think a lot of people have. You spend money on UPI all day, get SMS alerts you immediately forget, and then feel vaguely guilty at the end of the month without knowing where it all went. I wanted to design something that tracked your spending without you having to think about it.

Most expense tracking apps I looked at made the problem worse by adding more steps. Log in, find the add transaction button, fill in amount, category, account, save. By that point you have already forgotten the context. I wanted to design something where being financially aware required almost no effort at all.

¢
The core idea
Cent turns financial tracking from a manual task into an ambient behaviour. You spend, it notices, you tap once, it learns. Over time you do not even need to tap, it just knows.
02
Problem
Why existing fintech apps fail Indian users

The issue is not that people do not want to track their spending. Most people I spoke to genuinely wanted to be more aware of where their money was going. The issue is that every existing solution required effort at exactly the wrong moment, right after you have just paid for something and want to move on.

Manual effort barrier
Users must manually enter or verify transactions, leading to drop-offs within 5 to 7 days. The habit never forms.
📷
Fragmented financial view
Bank apps, UPI apps, and credit cards operate in silos. There is no single place to see the full picture.
🕐
Delayed awareness
Spending gets reviewed at month-end, when it is too late to change behaviour. Real-time feedback is missing.
📈
Low retention
Most users abandon expense tracking within a week. The apps treat tracking as a feature, not a habit.
03
Research
Behavioural and India-specific insights

I looked at two things in research: how people actually behave around money day-to-day, and what is specific to the Indian context that most expense apps are completely ignoring.

🧠
Recall drops fast
Users remember spending best immediately after a transaction. After 24 to 48 hours, recall accuracy drops significantly.
🕐
2 to 3 step limit
Users avoid apps that require more than 2 to 3 steps per action. Any more and the habit breaks.
🇮🇳
UPI dependence
India runs on UPI, but no expense app reliably syncs UPI transactions automatically. That is the gap.
📱
SMS as awareness layer
Users depend on SMS alerts for transaction awareness. This is a moment Cent can intercept rather than replace.
04
Personas
Two distinct user types with different needs

Two personas came out of the research pretty clearly. Both wanted the same outcome, financial clarity, but their relationship with effort was completely different. Riya would use the app if it was fast enough. Aman would only use it if it was invisible.

Riya, 26 — Conscious Spender
Uses UPI daily. Wants to control spending but forgets to track. Motivated but inconsistent.
Needs: quick, low-effort categorisation
Aman, 30 — Passive Tracker
Uses credit cards plus UPI. Wants insights but hates manual work. Checks finance apps rarely.
Needs: automation and smart summaries
05
User journey
Before and after Cent

Mapping the current journey made it obvious where the problem was. It was not that people did not care, it was that the window between spending and awareness was too wide. By the time someone opened a tracking app, the moment had passed and the motivation had gone with it.

Current journey (painful)
Step 1
Transaction
UPI or card payment made.
Step 2
SMS alert
Notification arrives, user reads it.
Step 3
Ignores
Does not open tracking app. Too much effort.
Step 4
Forgets
24 to 48 hours later, recall is gone.
Result
Month-end guilt
No action taken. Cycle repeats.
Cent journey (redesigned)
Step 1
Transaction
UPI or card payment made.
Step 2
Smart notification
Cent intercepts SMS, sends smart nudge.
Step 3
1-tap categorise
User taps once from the notification. Done.
Step 4
AI learns
App predicts future categories automatically.
Result
Real-time insights
Awareness builds passively. No effort needed.
06
How might we
Design opportunities from research

Each question came from a specific user behaviour pattern observed in research.

  • How might we reduce friction in expense tracking to near zero so users never feel the need to log manually?
  • How might we help users build financial awareness without requiring any active effort on their part?
  • How might we use real-time context, like the moment right after a transaction, to improve the accuracy of categorisation?
  • How might we design for habit formation rather than just feature adoption?
07
Features
Core and advanced, all tied to research

The feature list was built by going through every step in the painful journey and asking what would need to exist to make that step disappear. Anything that added effort was cut or pushed to advanced.

Core
Real-time transaction detection
Detects UPI, card, and SMS transactions the moment they happen. No manual input needed.
Core
1-tap notification categorisation
User categorises directly from the notification. The entire action takes 2 seconds.
Core
AI category prediction
Learns from user behaviour and starts predicting categories automatically within two weeks.
Core
Budget tracking with live updates
Budget progress updates in real time. Users see where they stand today, not at month-end.
Advanced
Behaviour-based insights
Patterns like spending 40% more on weekends that actually change behaviour. Not just raw totals.
Advanced
Bulk categorisation and undo
For missed transactions, batch categorise in one go. Undo any mistake without friction.
08
Information architecture
Four clear layers, one coherent system

The navigation was designed around the assumption that most users will only ever look at the home screen. If the home screen answers how am I doing today, most people do not need to go anywhere else. Everything else exists for users who want to go deeper.

🏠
Home
Spending overview
Recent transactions
Quick categorise
Budget status
Smart nudges
🅾
Budget
Category tracking
Live progress
Custom limits
Overspend alerts
Monthly reset
📊
Insights
Behavioural analytics
Spending patterns
Trend comparisons
AI summaries
Habit score
👤
Profile
Account linking
UPI and card setup
Custom categories
Notification prefs
Privacy settings
09
Impact
Estimated product outcomes, conceptual but grounded

These are estimated numbers based on the behavioural research, comparable fintech app data I found publicly, and what I would realistically expect given the specific friction points Cent removes.

Efficiency
85% less manual data entry
Categorisation: 30 sec down to 2 sec
📈
Engagement
3x daily interaction vs other apps
Notification response rate up 60 to 70%
🧠
Behaviour change
Tracking: 5 days to 20+ days per month
Overspending reduced by 18 to 25%
🤖
AI accuracy
60% on day one, 90%+ after two weeks
Improves passively as the user keeps using it
📉
Drop-off
Most budgeting apps lose around 70% of their users in the first week. That number is almost entirely driven by the effort of manual logging. Cent targets under 30% drop-off by removing that effort at the source.
Self-initiated concept project. Figma file and full case study available on request.
← back to work
01 / 05 · EY GDS India LLP

GenAI Product Suite

Role
UI/UX Designer
Type
AI Tool Design
Tools
Figma, FigJam, EY Design System
Company
EY GDS India LLP
Overview

This was probably the most challenging project I have worked on. I designed five AI tools from scratch including an interview guide generator, project charter builder, transaction analysis tool, value creation suite, and a diligence copilot. Each one had its own use case and user type, but they all shared the same core challenge: how do you design an interface where the main output is generated by an AI and the user needs to trust it, review it, and act on it?

Key design decisions
Impact
5
AI tools
from scratch
0 to 1
Full suite
shipped
Enterprise
Decision support
at scale
I worked on this at EY GDS India LLP. Confidential so I cannot share screens, but happy to walk through the process on a call.
← back to work
02 / 05 · EY GDS India LLP

Health Claims STP Agent

Role
UI/UX Designer
Type
AI Tool, Healthcare
Tools
Figma, EY Design System
Company
EY GDS India LLP
Overview

Health insurance claims reviewers process a huge number of cases every day, and the existing workflow made it hard to know where to focus. I was brought in to design a straight-through processing interface that surfaced the most urgent claims immediately. The core question I kept coming back to was: what does a reviewer need to see in the first three seconds?

Key design decisions
Impact
Faster
Claim review
time
Risk first
High priority
immediately visible
Less noise
Cleaner signal
to action ratio
I worked on this at EY GDS India LLP. Confidential so I cannot share screens, but happy to walk through the process on a call.
← back to work
03 / 05 · EY GDS India LLP

Sales Dashboard — CPG

Role
UI/UX Designer
Type
Analytics, Dashboard Design
Tools
Figma, EY Design System
Company
EY GDS India LLP
Overview

A performance tracking dashboard for a CPG sales team. What made this interesting was that two very different types of users, field reps and strategy managers, were looking at the same data but for completely different reasons. The design challenge was building one dashboard that served both without confusing either.

Key design decisions
Impact
2 users
One dashboard
that serves both
Cohort
Readable without
statistical training
Action
Every number
points somewhere
I worked on this at EY GDS India LLP. Confidential so I cannot share screens, but happy to walk through the process on a call.
← back to work
04 / 05 · EY GDS India LLP

Benchmarking Tool

Role
UI/UX Designer
Type
Enterprise, Data UX
Tools
Figma, EY Design System
Company
EY GDS India LLP
Overview

This one started with a clear problem: users could find data in the tool but they could not easily do anything with it. The benchmarking workflows also had data security requirements, so I added locked file states and access-controlled uploads as part of the redesign.

Key design decisions
Impact
Rebuilt IA
From scratch
based on real usage
Fewer steps
Data to
action faster
Secure
Access-controlled
uploads
I worked on this at EY GDS India LLP. Confidential so I cannot share screens, but happy to walk through the process on a call.
← back to work
05 / 05 · EY GDS India LLP

Margin & Portfolio Analysis Tool

Role
UI/UX Designer
Type
Finance, KPI Dashboards
Tools
Figma, EY Design System
Company
EY GDS India LLP
Overview

A cross-business-unit analysis tool used at strategy and CFO level. The data involved multiple business units, margin types, and overhead allocations, all of which needed to be comparable side by side without requiring constant view-switching.

Key design decisions
Impact
C-suite
Designed for
executive use
Multi-BU
Side by side
comparison
Cleaner
Than the
previous tables
I worked on this at EY GDS India LLP. Confidential so I cannot share screens, but happy to walk through the process on a call.