Freelance Branding Independent
Brand and art direction for food, culture and music clients - identity, space, packaging and social.
Co-Founder & Creative Director Maachis.art
Built the Instagram community from 0 to 58K followers. Led content design, cultural curation, collaborations and visual storytelling around Indian matchbox art.
Product Designer Zolve
Designed cross-border financial experiences, translating complex banking and credit journeys into clear user flows, prototypes and polished interfaces.
Creative Resident Bloomingreen Festival
Built the festival ticketing website and led UI/UX. Directed Bloom TV, a behind-the-scenes series documenting the making of the festival.
UI/UX Designer FlashPrep
Designed learning flows for an edtech platform built around bite-sized learning, spaced repetition and active recall.
BCA St Joseph's University
Bachelor of Computer Application
Banking for people with no credit history
Zolve is a cross-border neobank. It gives international students and professionals moving to the US a bank account and a credit card before they land - underwritten on their home-country credit file, with no SSN and no US credit history required.
Not everyone gets the offer
Defined the cohort first, with risk and credit - active users with real spend history, no delinquency, no fraud flags. The check runs before anything is shown, so nobody is pitched an upgrade they'd be declined for.
Then the whole journey was structured in low fidelity first - entry points, pitch, commit, re-issue, recovery.
Discovery, where the user already is
Eligibility shouldn't be something you go hunting for. The primary surface sits on the home screen, which every user opens first.
The profile menu carries the same offer for people who go looking deliberately.
Two nudges tied to real spend
The rewards section shows what the free card cost the user last month. The post-payment nudge fires at the one moment the user is already thinking about money.
Both are capped at three views, with a remind-me-later that resurfaces twice before going quiet for good.
The pitch had to read as a tier you qualified for
The pitch leads with concrete value - boosted cashback at the user's own top brands, Mastercard cover, credit worth real dollars - with FAQs inline to pre-empt the "what happens to my old card?" hesitation.
The same page serves both premium tiers, and reshapes for users who also qualify for a credit-limit increase by leading with the new limit instead.
Support agents were already converting people with one number
They read a user their own spend, then what that spend would have earned on a paid tier. Those calls converted, so I turned the script into a screen.
Two decisions between intent and commitment
How to pay, and agreeing to the terms. Annual is framed as the value choice - $99 a year against $9.99 a month on Signature - while monthly stays the low-commitment default so nobody is pushed.
The terms screen shows the user's own name embossed on the new card with their own name on it, so they can see what they are agreeing to before they confirm.
The riskiest second in the flow
The old card has just stopped working. Instead of a spinner, the processing screen animates the classic card physically transforming into the premium one, so the wait shows the thing being done rather than a spinner.
A virtual card is live immediately, so there is never a moment with no usable card. Delayed and failure states were designed alongside the happy path.
Winning the upgrade is only half of it - the user has to be caught
The flow looks simple because the complexity was resolved underneath it
Worked through with engineering, ops and legal, against Mastercard's re-issuance rules. The card re-issues on the same credit line so the score is untouched, pending bills carry to the next statement, and auto-pay moves across.
Surfaced where the spend already lives
Four surfaces, each tied to a moment the user is already thinking about the purchase.
The whole decision on one screen
Drag the slider to change tenure and the monthly number updates live. Tick a box to pay the fee from rewards instead of cash. Nothing hides behind a second screen until the user asks for it.
What changed after the first version
The first version buried the numbers people actually compare behind a "Show More." Someone weighing 3 months against 6 had to expand a drawer to see the trade-off - exactly the moment you don't want friction.
The second pass split the same information into two tabs and promoted three fields that had been hidden: the tenure, the exact instalment amount, and the first due date.
It has to stay honest about what's still owed
A plan can't be cancelled, but it can always be paid down faster
I reused the existing bill-pay screen so one keypad handles two intents: a partial prepayment that spreads across the remaining instalments, or a full foreclosure that clears and closes the plan. That keeps exits inside the payment and autopay logic that already exists.
