salahvi

What I work on

Organised by problem, not by project.

Projects age out; problem spaces don't. These four are the domains I've spent real time inside — what each one actually is, and what it taught me.

01

Cards

Payment gateway SDKs

A payment SDK is a product with exactly one user who never chose it: the engineer integrating it under a deadline. The real problems aren't cryptographic — they're integration time, binary size, and error messages that say what actually went wrong instead of ERR_UNKNOWN. Every ambiguous field in the docs becomes a support ticket, then a failed payment, then a lost sale.

Working here taught me that developer experience is a payments metric. The SDK's job is to disappear — to make the correct integration the easy one, and the insecure one hard to write by accident.

02

Rails

UPI

UPI is built on rails you don't own. As a third-party app (TPAP) operating through a sponsor bank's PSP, you inherit a spec, a switch, and a set of guarantees you can't change — and you build a product on top that has to feel instant and forgiving anyway. The plugin model is the whole challenge in miniature: your surface, the network's semantics.

The lesson is about designing on borrowed ground — where "it works on our side" isn't the bar, because the parts that fail are the parts you don't control, and the user blames you either way.

03

Authentication & certification

EMV 3DS and certification

3DS answers one question — is the cardholder who they say they are — and shipping it under a certification body reshapes how you build. The specification is the contract; conformance is graded by someone outside your company; and "we'll fix it next sprint" isn't available when the fix has to pass a test suite you don't administer. The roadmap bends around approval windows, not just customer requests.

This is the part of the CV hardest to fake, and the part I find most interesting. It's also the subject of the interactive explainer — the honest way to show protocol depth is to let you walk the protocol yourself.

04

Earlier — 2014–2020

Design and frontend

Before the platform work, there was the surface. Six years of design and frontend engineering — interfaces, design systems, and the craft of making a screen legible under pressure. I treat this as foundation, not baggage: it's why I care what a payment feels like, and why the 3DS explainer shows the cardholder's screen next to the message flow.

A few pieces from that chapter, closed on purpose:

  • Coinome crypto exchange

    Trading and wallet interfaces for a cryptocurrency exchange — the most demanding thing I built in that period, and squarely on the money-movement thread that runs through all of this.

  • Western Union remittance

    Front-end work in the cross-border remittance space — money moving between people and countries, the interface end of the same problem I'd later work on from underneath.

  • Paymetry payments

    A payments-adjacent product — an early turn toward the domain that would become the whole career.

  • Design-system & interface work craft

    A body of visual and frontend work from the period — kept as evidence that the design half is real, not decoration.