Skip to content

Uttam Paswan

Lead Product Designer

Back
  • PayRam
  • Web3
  • Payments
Figma opens in a new tab

In testing, customers understood the payment, not the fee. After Coins, all of them paid on their own.

Making stablecoin payments for people who are not crypto experts.

Problem:

Users read the fee as spending. In February, all 20 testers were confused by the fee wording.

What I did:

I proposed PayRam Coins, so the fee reads as earned back. I also unified checkout into one screen.

Results:

In April, all 20 testers paid on their own.

In beta, 200 of about 500 users paid.


Role

Sole designer

Jan - Apr 2026

Team

  • PM
  • CEO
  • CTO
  • Eng

Tags

Design strategy, UI, UX, User Research.

Web checkout & mobile app

PayRam lets merchants take stablecoin payments on their own servers. Customers pay through the app I designed. PayRam reached $100M+ in on-chain volume in 2025 and 2026. That is a team result, not this app’s.

The Problem

The idea was simple: a fee that felt like a reward, and a flow people could finish. Merchants would fund up to 2 to 5% off, the same range as the onramp fee we expected from a partner.


  • All 20 testers were confused by the fee refund, cashback and offer wording.
  • 8 asked why they had to add funds first.
  • 6 questioned the adding-funds step on the Onramper side.
  • 4 finished paying but found it heavy to follow.

Sample: 10 beginners, 5 intermediate, 5 power users. One-on-one calls. Feb 2026

If customers cannot finish paying, the merchant loses the sale.

Role

I was the founding designer at PayRam and the sole designer on this app, January to April 2026. The core team was six, with legal and marketing. I designed the flows, screens and Coins interaction. Three constraints shaped the design:


  • Self-custody: PayRam never holds the funds. Every step needs the user's confirmation.
  • A third-party onramp for card-to-crypto, with limited partner data.
  • Safari, Firefox and in-app browsers, some of which block redirects.

Decisions


Chose a partner wallet over building our own

Why:
We were building our own self-custody wallet. In February we found Privy, which already did wallet creation and analytics.
What it shipped:
Wallet creation and Google sign-in through Privy.
What it cost:
We now depend on a partner for wallet creation.
Proof:
Privy saved an estimated one week of design and two weeks of engineering.

Chose a small reward layer over a full game

Why:
Testers did not respond to fee credit, cashback or offers. I proposed Coins, and Karan and Siddartha agreed to explore it. I explored gamification from 1 to 10 in Figma Make, then shipped a light layer.
What it shipped:
Fees return as PayRam Coins, one Coin per dollar of fee. A separate Coins space with a ledger. A prompt for users with no Coins, with a nudge to add funds and start earning. A full payment breakdown showing Coins deducted and earned. Light Lottie animations in our branding.
What it cost:
Merchants fund the Coins, like a card fee. Each merchant sets the share of an order payable in Coins and a maximum per order. Coins cannot be transferred.
Proof:
Testers understood the payment but not the fee. Most grasped the idea only after Coins and the light gamification. All 20 then paid on their own.

Chose context before the onramp over a direct hand-off

Why:
Eight testers asked why they had to add funds first. Landing straight on the onramp made it worse. It showed no recipient, no amount and no reason, and its different branding read as untrustworthy. So our screen explains the shortfall first.
What it shipped:
An error on the amount field with the shortfall, and an Add funds action beside it.
What it cost:
More copy on a failure screen, at the moment of frustration.
Proof:
The April test came after Coins.

Solution

Customers pay on one screen. If the balance covers the amount, they pay directly. If not, they add funds first. The fee returns as Coins.

The payment window opens over the dimmed merchant page, so the customer stays in the same context.
Coins applied and the Coins earned appear in the payment breakdown, so the fee reads as earning.
One send screen handles balance cases: an error on the field, a nudge to add funds, or send less and restore the amount.

Results

  • 20 of 20 paid on their own after Coins

    Onboarded testers, early April 2026.

  • 40% of about 500 beta users completed a payment

    Clients’ beta release

  • 1 screen covers every balance case

    Pay directly if the balance covers it, otherwise add funds first.

What I learned

Reflections

  • A simple idea hid a lot of unknowns across use cases.
  • I designed the flow before the partner integration was proven. Next time I would push for a working sandbox first.
  • Design could not change what the onramp charged. We expected 2 to 5%, and it ran 8 to 15%.
  • The open question is what makes someone add funds before they need to.

Shout-outs.

  • KaranProduct
  • SiddarthaCo-founder
  • SamirCTO
  • ManishEngineering

Questions I actually get

Was it shipped?

A limited beta in April 2026. It is not live.

What did you own?

I was the only designer on the app, from January to April 2026. I owned the UX, wrote the product requirements and designed every flow.

How did you work with engineering?

Daily stand-ups with the PM, tech leads and co-founders, and a weekly meeting with the whole team. We tracked work in Jira and talked on Slack and Telegram. Karan, our PM, was in the US, Siddartha in the UAE and the rest of us in India. We met in the UAE every quarter. Because the product was technical, I sometimes called Manish and Gulshan to understand the logic behind a requirement. I also joined customer and sales calls to hear what people asked.

Is 20 testers enough?

Enough to see a pattern. In February all 20 were confused by the fee wording, and in early April 20 of 20 paid on their own after Coins. It does not prove a rate. For scale, about 200 of about 500 beta users completed a payment.

Working on payments or wallet UX?

Email me, and I'll walk you through this flow in Figma.