I'm Shayan Rasheed, a product manager who turns ambiguous, real-world problems into shipped outcomes. I diagnose what's broken, design the fix, and increasingly build it myself with AI. Deep in payments, built to work anywhere the problem is messy.
Before and after.
Our Raast payment success was stuck around 31%. When I dug into the data, most transactions weren't failing. They were expiring, because people couldn't find where to approve the payment inside their own banking app. I shipped one page that showed each user exactly what to tap, and success climbed to 65% across 14 banks.
On Pakistan's biggest travel site, the real price kept arriving too late. A card fee at the last step, foreign-transaction tax on the statement days later, and a call center that kept hearing one word: scammed. I ran three controlled A/B tests that moved the true price forward through the funnel. Method mix, mid-funnel progression, and repeat booking all went up, and fee complaints dropped 70%.
1Bill had no extra fees, so people kept choosing it, then couldn't actually pay. Most transactions sat in pending and expired, because every bank app had its own way to finish the transfer and nobody, not even our support team, knew all of them. I built a tutorial system that showed each user how to pay with their specific bank. Stage-one drop-off fell from 80% to 16%.
Safepay's card tokenization worked, but using it took nine steps across six token types, far too raw to hand a merchant. I designed one merchant endpoint that hides all of that behind a single call, so returning customers at Yango and Temu pay card-on-file. It unlocked $251K+ in new monthly volume.
About 22% of our Cybersource card payments were stuck in pending, never reaching success or failure. My dev team didn't want to dig into it, so I did it myself. I traced the data, listened to support calls, and watched session recordings until I found it: the Pay Now button never disabled. People tapped it around ten times, each tap firing a duplicate request that broke 3DS. Disabling it on first tap lifted success about 10%.
I'm a product manager. I've shipped payments in Pakistan, Bangladesh, Nepal and Egypt, markets where the flow on paper and the flow in someone's hands rarely match. Every fix I'm proud of came out of a room where someone knew more than me: an engineer, a compliance officer, a support agent halfway through a call. I've learned to treat that as the raw material.
My job is to work out what to build. I read the data, listen to support calls, watch session recordings, and try the broken flow myself until I find the real reason it isn't working. It's usually smaller and stranger than anyone expected. Then I design the smallest change that fixes it and get it built.
Payments has too many moving parts for one person to hold them all, so I try to build rooms where saying "I don't know yet" counts as progress. When a Spotify launch was about to slip because our sandbox couldn't be ready in time, admitting that early gave us room to act: I built a mock of the bKash API overnight, and Boku integrated against it, a first for their company. The slot held, and 260K transactions have run through it since. When speed matters most I still build the first working version myself, with AI carrying more of that every year.
Operators, platforms, and marketplaces I’ve built solutions for. Product, payments, growth, and AI work, shipped into live systems across Pakistan, Bangladesh, Nepal, and Egypt.