Juspay HyperUPI & TPAP
Designing UPI that scales - onboarding, payments, and the system underneath.
Role
Product Design Lead
Duration
6 months to ship but still going on...
125 Mn+
UPI txn/day
50k+
UPI txn/sec
So we looked at the funnel step by step.
The answer was uncomfortable: the biggest losses weren’t at the hard steps. They were at the permission dialog and the bank-linking screen - 26% and 36% of users gone, before anyone had entered a single digit of a PIN.

We rebuilt the journey from the drop points outward - pre-empting permissions with context, auto-detecting the user's bank, and skipping VPA creation entirely.
Repeat payments went from 4 inputs and ~30 seconds to 2 inputs and ~6 seconds.
But the bigger unlock was underneath: instead of designing screens, I designed the config layer they render from. MasterConfig gave devs a single point of control over theme, components, icons and layouts, so a merchant's entire brand expression became a config change rather than a code change.
UPI Studio turned those tokens into a no-code surface - which meant a fully branded pitch flow could be built for a prospect in an afternoon.
The strange part of the brief: this isn't an app. It's a set of screens that get dropped inside someone else's wallet, wearing their colors, running on infrastructure the user will never see and would not enjoy having explained to them.
You can't onboard someone into infrastructure. You can only onboard them into a moment.
18%
Upward in successful user onboarding
40+
Merchant themes shipped without an SDK release
<2 hrs
To build a fully branded pitch flow

This is a confidential project and hence can only share more information over call