tap to skip

I love designing data heavy products, with a dash of AI and a little vibe coding fun.

Current obsession
Vibe coding
Tools
Claude, Replit, Cursor, OpenAI, Figma
Interests
Mood based web apps, data heavy projects, AI products

Built by me

My own product, from first idea to real users.

Bitewize recipe results served on plates

Bitewize

Solo built · AI

An AI food discovery web app. A 22 person survey showed health goals and cravings pull equally hard, so it became a cravings first, tap to add flow with a warm, encouraging tone.

User researchLovableCursor
15+users in the first month
See the research

Case studies

Each one follows the same arc: the objective, what I did, and what changed.

A box of mini apps

Little web apps I vibe coded on Claude, mostly for myself. Tap a sticker to try one.

Other work

Freelance projects I designed outside a design team.

Freelance and personal projects

ATC logistics dashboard Deliveries, transit time and trucks by state
Atlys visa app screens
Atlys visa app Visa booking with a delivery date on every card
Management dashboard screens
Management dashboard Sales, material requests and approvals
Chart components screens
Chart components Stacked, grouped and radial charts

Kind words

From the founder, product managers and developer I've built with.

Prabhav SharmaCEO, Zintellix
Manali has good product sense. She understands the latest trends in design, especially when it comes to dashboard designs. She stays dedicated and committed to her projects.
Meghana VunnamPM, Smytten
Manali is a very enthusiastic and passionate designer. She has always been very reliable and flexible with learning new things and delivering top notch product designs.
Rahul PhillipSenior PM, Smytten
Manali is an outstanding UX designer whose work consistently exceeds expectations. Her talent for building products from scratch, improving existing systems, and creating user centric designs has greatly benefited our team.
Mayur ShindeLead Developer
Working with Manali makes development faster and significantly smoother. She understands technical constraints instantly and always comes back with developer friendly, clean, and scalable solutions. Her handoffs are clear, logical, and easy to implement, the kind that save hours of rework.

When I'm not designing

You'll probably find me travelling, in a pool or painting. If you do any of these too, tell me. That's my favourite way to start a conversation.

Take a break on the farm

Five small chores, about a minute. Feed the cows, bring the chickens home, take the horse for a lap and spoil the dog and the cat.

Tap or drag things to help out. Stuck? Tap the i for help.

About

Manali Gupta smiling, seated at a wooden table

I'm Manali, a product designer from Jaipur. I joined Smytten Pulse as an intern and left as its Senior UX/UI Designer, designing the dashboards that Pulse's clients, including brand teams at HUL, Colgate and Dabur, used to make decisions.

Since then I have designed an AI shopping browser at Sellergize, freelanced, and built my own products with AI tools. I care about making complicated things feel calm.

  1. Jan 2025 to Mar 2026UX Designer, Sellergize WTSZifup, AI shopping browser. Mentored designers on AI assisted workflows
  2. Nov 2024 to Jan 2025Freelance, ZintellixEnd to end product design for a 0 to 1 MVP
  3. Oct 2022 to Nov 2024Intern to Senior UX/UI Designer, Smytten PulseB2B consumer research platform. Built its design system from scratch
  4. EducationB.Des, UX/UI, Jagannath University (ImaginXP)9.0 CGPA, ranked 2nd in batch

Giving back

  1. Since Jan 2021Teacher, UphaarI teach kids English and the basics of art, and show the ones who love sketching that art and design can be a real future for them.
  2. Feb 2020Art Coordinator, ContreeWith an NGO team of 14, we painted a government school's walls with facts on evaporation, flora and fauna, world leaders and athletes. I decided which design went on which wall, after we talked with the kids about their goals and how they learn.
View full resume
All work
Case studyB2B SaaSConsumer researchWeb platform

Smytten Pulse

A self serve research platform that lets FMCG brands ask real consumers and read the answers as live dashboards.

Enterprise dashboardsSurvey builderData visualisation
Role
Senior UX/UI DesignerIntern to Associate to Senior
Project timeline
3 monthsPart of my two years at Smytten
Platform used by
HUL, Colgate, DaburSmytten Pulse's enterprise FMCG clients
Recognition
IAMAI 2024Best Technology Innovation
Objective

Research without the wait

Brand teams needed answers from real consumers fast enough to act on them. Pulse had the audience: Smytten's high intent shoppers. The job was to turn that audience into a product brand managers could use on their own, from choosing a study to reading what it found.

Before Pulse, brand research relied heavily on fragmented workflows: extracting data from databases, managing spreadsheets, and manually compiling reports. This made it harder for brands to track awareness, recall, competitor performance, and changing consumer perceptions efficiently.

I worked with product managers and engineers, using stakeholder discussions, product workshops, and analytics to understand the existing workflow and translate research needs into an integrated survey and insights experience. Brand Track also enabled brands to schedule recurring studies and track performance over time.

Action

From study to spend

Research hub

Make the menu legible

Four study types, each with its own colour, credits and starting templates, so a brand manager can tell product testing from a brand track at a glance.

Research hub with consumer research, quick surveys, post sampling feedback and brand track
Survey builder

Build with the respondent in view

Mandatory and optional metrics are set first, then questions are edited beside a live phone preview of what the shopper will see.

Survey builder with metric selection and mobile preview
Results

Read it, compare it, pin it

Audience persona, clustered histograms with AI insights, and a compare view that puts two research waves side by side.

Brand track results compared side by side
Study list

Every study, one status

Each survey shows whether it is created, reviewing, scheduled or active, with the next action, Approve or Launch, right beside the status. Pinned views sit above the list.

Brand track results list with survey status and approve and launch actions
Finance admin

Money moves get a second look

An internal tool for the finance team to add balance, set credit limits and download invoices per brand. Every change opens a confirmation showing the new balance before it is applied.

Finance admin with balance, credit limit and downloadsConfirmation dialog showing new balance before adding funds
Underneath it all

I built Pulse's design system from scratch. It became the shared foundation the research hub, survey builder, results dashboards and internal tools were all built on.

Outcome

What changed

25%reduction in time to insight
18%increase in offer conversion
IAMAI2024 Best Technology Innovation award
What I learned

I learned that B2B product design is as much about simplifying complex workflows and making data actionable as it is about usability. Designing around how teams actually work helped me make the product more scalable and useful.

Next case study

Zifup

All work
Case studyConsumer AIShopping browserMobile

Zifup

An AI shopping browser that finds the best coupon and applies it to your cart, with every marketplace order in one place.

AI nativeE-commerceUser research
Zifup home screen explorations: savings counter, search and lowest prices
Role
UX DesignerSellergize WTS, also mentored designers on AI workflows
Timeline
Jan 2025 to Mar 2026Through launch and iteration
Scope
Research to UIJourneys, flows, persona, screens
Status
Live on mobileiOS and Android
Objective

Save without the hunt

Shoppers lose time searching for coupon codes that may not work, then lose track of orders spread across several marketplaces. Zifup set out to handle both: find the saving automatically, and keep every order in one view.

"When I open Zifup, I want to know I'm getting the best price without spending hours comparing tabs, searching for coupon codes, or worrying about fake reviews."Emily, the primary persona we designed for
Action

Map it, then make it quiet

Research

Follow the order, not the app

I mapped what people do after they buy: cancelling, changing a delivery date, waiting for installation, chasing a warranty. Each step showed where they fall back on customer care. A persona and a flow chart for the browser came out of it.

Whiteboard plan for Zifup phase one: timeline to go live, onboarding, coupons, price comparison and tracking, reviews, with a screen sketchJourney map table, browser flow chart and persona
Phase one planned on the whiteboard, then mapped in detail.
Coupons

Best price, without leaving the page

On a product page the browser shrinks the store to 70% height and slides up an overlay comparing prices, coupons and vouchers. The best code is applied for you. Home keeps a running total of what Zifup has saved.

Home explorations with savings counter and lowest prices with coupon applied
Flexbook

Every order, one book

Flexbook syncs orders from stores like Amazon, Flipkart, Myntra and Ajio into one timeline grouped by month. Filters by status and store, and a sync manager to choose which stores connect.

Flexbook explorations: monthly order grid, order detail, filters, sync manager
Three rounds of Flexbook explorations, from wireframe to final colour.
Outcome

What changed

I tested prototypes with users before launch and kept iterating after it shipped.

TestingUsers found Zifup worth it because it made finding and applying discounts easier, but they didn't buy the idea of replacing their everyday browser with it.
IterationWe repositioned Zifup from a browser to an AI shopping assistant. Instead of waiting on the user's search, it now guides them from the start of the journey to discover products.
ShippedLive on iOS and Android, with auto applied coupons and Flexbook for orders from every marketplace.
What I learned

Users loved the savings but not a new browser. I learned to separate what people value from the shape we deliver it in, and to let that decide how the product shows up.

Some screens and details from this project are under NDA. I'm happy to walk through more of the process in a conversation.

Next case study

Zintellix

All work
Case studyFreelanceB2B AIAnalytics

Zintellix

An AI data platform that brings a brand's tools into one dashboard, with Zin, a data assistant, a click away.

DashboardsIntegrationsAI assistant
Zintellix dashboard: KPI cards, top selling products, competitors, product activities and inventory
Role
Freelance Product DesignerEnd to end, 0 to 1 MVP
Timeline
3 monthsNov 2024 to Jan 2025
Surfaces
Web appDashboard, connectors, admin
Client
Early stage startupProduct built from the ground up
Objective

One place to look

Sales sit in one tool, ads in another, product work in a third. Zintellix connects them so a team can see sales, margins, competitors and stock on one screen, and ask Zin when they need more.

The client was shaping an early stage product from the ground up. My brief was to translate the product requirements into a clear information architecture, research competing solutions, and design the dashboard experience, from product structure and sitemap through to high fidelity screens, over about three months.

Action

Connect, then compare

Data connectors

Grouped by job, not by brand

Google Analytics, Tableau, Power BI, GitHub, ClickUp and the ad platforms are sorted into analysis, product and marketing tools, so setup reads like a checklist.

Data integration screen with analysis, product and marketing tool cards
Dashboard

Comparison is the default

Every block compares something: this week against last, a brand against its competitors, margin against profit and growth. Inventory sits below for detail.

Dashboard with weekly comparison chart, competitors and product activities
Outcome

What changed

I delivered the foundational product experience for the MVP: the sitemap, information architecture, competitive analysis, and dashboard designs. The work gave the product a clearer structure and a design foundation for implementation.

SitemapInformation architectureCompetitive analysisDashboard UXHigh fidelity UI
Next case study

Smytten Pulse

Built by me
Personal projectConsumer AISolo built

Bitewize

An AI food discovery app for the moment you can't decide what to eat. I researched it, designed it and built it myself.

User researchLovableCursorTry the live app
Role
Solo designer and builder
Research
22 survey responsesFood and lifestyle habits
Users
15+ in month one
Objective

The daily what to eat

Deciding what to eat is a small problem that shows up every day. Before designing anything I ran a survey on food decisions, habits and what would actually help.

82%struggle to decide what to eat at least a few times a week
73%say health goals drive their choices, and 73% say cravings do
55%want simpler healthy recipes, the top ask
Research

Health and cravings tie

What influences your food choices most?

22 responses, pick up to 3
  • Health goals73%
  • Taste, cravings73%
  • Time, energy59%
  • Convenience45%
  • Mood, stress41%
  • Budget36%

What would help you eat well more consistently?

22 responses, pick up to 3
  • Simpler healthy recipes55%
  • Better routine50%
  • Easier meal planning45%
  • Workout based tips27%
  • Grocery planning27%
  • Personal suggestions23%
"It's complicated. Switching from healthy to junk."Survey respondent
Raw responses

The survey, as it came in

Four sections: about you, food decisions, lifestyle and wellness, and product opportunity. The open answers kept coming back to late night cravings, skipped meals and stress.

Action

Start from the craving

The pivot

From ingredients to cravings

The first concept asked people to drag in the ingredients they had. The research said people start from what they feel like eating, so the flow became cravings first: tap what you want, get a healthier way to have it, in a tone that encourages instead of judging.

Craving picker with six bright tiles: fresh, savory, spicy, sweet, comfort and crunchy
Results

Dishes that cheer you up

Recipes arrive served on plates, with filters for craving, cuisine and diet in one row above them. Changing your mind takes a single tap.

Recipe

From craving to cooking

Each recipe opens with ingredients and quantities, then short numbered steps, and a link to watch it being made.

Outcome

Real users, early

15+users in the first month
22people surveyed before the first screen
Next to validateWhether people can get from a vague craving to a dish they want with less decision effort.
Back to case studies

Smytten Pulse