Yash Verma Résumé↓

I turn declines into approvals.

Seven years as a seasoned card product manager, inside the 300 milliseconds between a tap, or a click on Pay, and an approval. I build the checkout, 3DS, routing, authorization and settlement rails on platforms processing $10B+ a year.

Product Manager II · Payment Gateway · Wibmo, a PayU company

Follow a transaction↓ Get in touch
+4% success rate · routing engine◆ −5% authorization errors◆ −15% checkout bounce rate◆ Refunds 3 days → 15 min◆ Onboarding 30 merchants 50 → 2.6 min◆ $60M+ monthly on-demand settlement GMV◆ 8 new international banks◆ PMOS · AI agents for product work◆
scroll

Prologue

A card tapped at a counter. A Pay button clicked on a checkout page. Either way, the authorization behind it takes about 300 milliseconds — and decides whether a sale happens. Most people never think about them. I’ve spent 7+ years inside them — making sure the answer comes back yes.

Two front doors

Tapped at a counter or paid on a checkout — the same 300 ms.

In storeCard-present · tap to pay t+000ms
  • EMV contactless · NFC
  • Tap on Phone · SoftPOS
  • Device wallets
  • CVM limits · PIN
  • Terminal · MID / TID
OnlineCard-not-present · checkout t+000ms
  • Hosted checkout · iFrame
  • EMVCo 3DS 2.x
  • Network tokens · card-on-file
  • Native OTP
  • Alt ID guest checkout
  • UA-aware flows
One rail
  • ISO 8583 · ISO 20022
  • Card schemes
  • ON-US / OFF-US routing
  • Smart routing

Two front doors, one rail behind them — and the rail is what I build. Follow it ↓

Success rate = payments completed ÷ payments attempted. Every result below names the company it was achieved at.

01 Authenticate Online3DS · pre-auth

Prove it’s really you — without the friction.

Online, the card isn’t in the room, so the issuer asks for proof before it will even look at the payment. Every extra screen here is a buyer who might not come back. My job is to make that check pass on the first try.

  • +3% Authentication success, one layerWibmoMerged the PayU and Wibmo EMVCo 3DS stacks into one orchestration layer, so every issuer gets the same clean, spec-compliant request.
  • 1 mo Early on EMVCo 3DS 2.3Recertified with two major card networks a month ahead of plan — 20% faster to market.
  • +3% Success rate, fewer stepsEaseBuzzAlt ID guest checkout, network tokens with a Token Management System, and Native OTP that keeps the buyer on the merchant’s page instead of a bank redirect.

EMVCo 3DS 2.xTokenizationNative OTPAlt IDNetwork passkeyClick to Pay

2FA

Why it mattersIn India, nearly every domestic online card payment needs a second factor — so this step sits on almost every checkout.

02 Route In storeOnlinet+40ms

Send every payment down the road most likely to say yes.

The same card can reach its issuer through several acquiring banks — and in any given hour, one of them is quietly declining more than the rest. Routing reads those signals live, payment by payment.

  • +4% Success rate, across acquirersWibmoA routing engine that scores every acquiring bank on live approvals and moves traffic away the moment one starts to slip.
  • +2% Success rate, card by cardEaseBuzzA BIN Management System that knows each card’s issuer, network and type, so it takes the lowest-cost acquirer route that still approves.
  • $34K+/mo Revenue from on-us routingEaseBuzzWhen the card’s issuer is also the acquirer, the payment stays on-us: lower cost, fewer hops.≈ $400K a year, from one routing rule

Smart routingBIN managementPayment orchestration

$5M

Why it mattersWhat one extra point of approval rate is worth on $500M a year in payments.

03 Authorize In storeOnlinet+180ms

Speak the networks’ language — fluently.

Every decline comes back as a two-digit code in an ISO 8583 response. 05, 51 and 91 mean very different things, and each needs a different fix. Reading them closely is how declines turn into approvals.

  • −5% Authorization errorsWibmoWorked response-code patterns with card networks and issuing banks to separate fixable technical declines (timeouts, format errors, issuer unavailable) from genuine ones — approvals up 2.5%.
  • 100% Mandate readinessWibmoEvery card-network ISO 8583 mandate ready on time across authorization and clearing, plus ISO 20022 on the bank-settlement side.Zero missed network deadlines
  • 8 New international banksWibmoOpened acceptance for AMEX, Diners/Discover, JCB, UnionPay and Saudi Arabia’s MADA — each scheme with its own message spec and certification.

ISO 8583ISO 20022Card schemesRBI + card network mandates

300 ms

Why it mattersRoughly how long card networks give an issuer to approve or decline. The whole decision happens inside it.

04 Approve Onlinet+300ms

Then make the yes feel effortless.

An approval only counts if the buyer is still there to see it. A lot of checkout drop-off isn’t the bank saying no — it’s a slow page, a broken redirect, or a payment app that never opened.

  • −15% Checkout bounce rateEaseBuzzRebuilt the hosted checkout page and the iFrame that merchants embed in their own sites.
  • +2% Success rate on mobileEaseBuzzUser-agent-aware flows and iOS deep-linking open the right bank or UPI app every time, with Native OTP instead of a redirect.
  • 4,000+ Active merchants, from 50EaseBuzzRevamped the merchant app: from 50 active merchants to 4,000+ in six months.80× adoption

Checkout pageiFrameMobile web & appsDeep-linking

7 in 10

Why it mattersOnline carts are abandoned before purchase, on average (Baymard Institute).

05 Settle In storeOnlineT+1 → T+2

Approved is a promise. Settled is the money.

After the beep comes the back office — clearing, staging, settlement and recon over T+1 and T+2, then refunds and disputes. It’s the part merchants judge you on.

  • $60M+ On-demand settlement, monthlyEaseBuzzNBFCs, lenders and real-estate firms get paid the same day instead of waiting for T+1 or T+2. The gateway fronts the money, so every payout is a credit decision.≈ $2M paid out every day
  • 3 days → 15 min Refund turnaroundEaseBuzzPlus instant UPI refunds, live for 3,000+ merchants.≈ 290× faster
  • $560K+ Revenue leakage stoppedEaseBuzzA dispute resolution system and AI-assisted compliance suite, at 88% adoption.

ClearingStagingSettlementReconRefundsDisputes

  1. After the beepT+0 → T+2
  2. T+0 Batch cut-off The day’s approved payments close into a batch at end of day.
  3. T+1 Clearing The networks’ clearing files confirm what was really captured — amounts, fees and interchange.
  4. T+1 Staging Files land in staging, get validated, de-duplicated and matched to their authorizations before money moves.
  5. T+1 / T+2 Settlement Funds move issuer → network → acquirer, then out to the merchant’s bank account.
  6. T+2 Recon Every transaction matched across bank, network and merchant ledgers; the breaks become refunds or disputes.
5–7 days

Why it mattersHow long a normal card refund takes to reach the customer in India. Instant refunds are the exception, not the rule.

06 Scale In storeOnline× every merchant

Then do it for everyone, every day.

The last step isn’t technical. It’s turning one good payment into a product that banks, merchants and sales teams can run without me in the room.

  • 5,000+ Merchants onboarded a monthWibmoThrough an API-based onboarding solution, plus a servicing portal now used by 10+ partner banks.≈ one new merchant every 9 minutes
  • $900M+ Government GMV, cumulativeEaseBuzzElectricity bills, state treasury receipts and tax payments — BSES, e-GRAS, CBDT and CBIC.
  • +20% Domestic business growthWibmoSolutioned and priced enterprise deals worth $500M+, and helped win 15+ clients across EMEA and APAC.

Onboarding APIsBank portal · 10+ banksEnterprise

KYC

Why it mattersEvery new merchant needs full KYC and due diligence before its first payment. At thousands a month, only automation keeps up.

Before the rails

I learned growth before I learned ISO 8583.

That’s why I build payment products the way a growth team would: obsessed with the funnel, measured by conversion. It’s also where I built my first payment layer.

Lead Product Manager · 2019–2021 · early-stage startup

EventBeep

Built a student community across five interest areas and rebuilt the funnel around it.

+67%
App downloads
9.5%
Conversion with a “Show interest” flow — about 1 in 10
26%
DAU / MAU within six months — about 8 active days a month

Also built · the payment layer

Orchestration middleware between the app and multiple payment aggregators — one integration for the product, with every payment routed to the right aggregator.

Product Growth Manager · 2021–2022

Unschool

Ran growth across B2C subscription and B2B, owning paid acquisition across search, display, performance and email.

35%→15%
Cost of acquisition
2 mo
Early on a digital roadmap, at 15% lower spend

Also built · the payment layer

Orchestration middleware between the app and multiple payment aggregators — one integration for the product, with every payment routed to the right aggregator.

Career · 2019 → now

Seven years. Every stop led to the rails.

2018

Jul 2018 – May 2021

B.Com (Hons), Accountancy & Taxation

D.Y. Patil International University, Pune

  • Ledgers, reconciliation and tax — the back office of money, before I ever touched a card
2019–21

Oct 2019 – Dec 2021 · alongside my B.Com

Lead Product Manager

EventBeep · student community

  • Built a community across five interest areas; DAU/MAU reached 26% in six months
  • +67% app downloads; a “Show interest” flow converting at 9.5%
  • Built my first payment layer — middleware across multiple payment aggregators
2021–22

Dec 2021 – Jun 2022 · Hyderabad

Product Growth Manager

Unschool · edtech, B2C + B2B

  • Cut cost of acquisition from 35% to 15%
  • Shipped a digital roadmap two months early, at 15% lower spend
  • Routed course and subscription payments across multiple aggregators
2022–24

Jul 2022 – Apr 2024

MBA, Product Management & Business Analytics

Sri Balaji University (BIMM), Pune

  • Product strategy, analytics and SQL — then straight into payments, mid-degree
2023–25

May 2023 – Jan 2025 · Pune

Associate Product Manager, Payment Aggregation

EaseBuzz · payment aggregator, 150,000+ merchants

  • Tokenization, Alt ID and Native OTP: +3% success rate
  • BIN and on-us routing; on-demand settlement at $60M+ a month
  • Government integrations — $900M+ in tax and utility GMV
2025 →

Jan 2025 – now · Pune

Product Manager II, Payment Gateway

Wibmo, a PayU company

  • A routing engine across acquiring banks: +4% success rate
  • Unified the PayU and Wibmo 3DS stacks; EMVCo 2.3 recertified a month early
  • ISO 8583 / 20022 mandates and 8 new international banks
Next

Your team?

Card payments, gateways, routing, 3DS — let’s build the next yes.

Let’s talk →

What I’d build next

Three bets on where card payments are heading.

01 · Authentication

OTP becomes the fallback, not the default.

RBI’s 2026 authentication directions require two factors but no longer insist on SMS OTP. Device-bound factors — biometrics, passkeys, in-app approval — can take the step that loses buyers today.

April 2026RBI’s two-factor directions take effect

02 · Credentials

The token is the card.

Since October 2022, Indian merchants can’t store card numbers, so the token became the credential. The next gains come from treating token lifecycle as a product: provisioning, updates and token-based authentication.

560M+card-on-file tokens in India by October 2023

03 · Rails

Cards and UPI share one checkout.

Credit cards now run on UPI as well as on card rails. Routing has to choose across rails, not just across acquiring banks — the same problem I solve today, one level up.

~38%of India’s credit-card transactions by volume now run on UPI

Off the rails · AI

I build routers for payments. Then I built one for product work.

PMOS is my product management operating system: nine AI agents, one for every seat at the product table, orchestrated to work as one. Underneath it runs Agent HQ — describe the output you need, and it calls the right agent for the job, or creates a new one.

Incoming request · illustrative demo — pick one, or just watch it run

Output · Waiting

PMOS

A product management operating system

Nine specialist agents that hand work to each other like a real product team: research, PRD, a Technical PM who writes the spec and acceptance criteria, a Developer who builds to them, a QA Engineer who tests against them — then security, analysis and data science.

Agent HQ

Agents, on demand

An architecture for spinning up an agent for any use case. Define the output you want; HQ routes the request to the agent built for it — and creates that agent if it doesn’t exist yet.

The through-line

Same instinct, new rails

A routing engine sends every payment down the path most likely to succeed. Agent HQ does the same for work: the right request, to the right agent, first time.

Toolkit

My skills, as an authorization message.

Every ISO 8583 message carries a bitmap saying which fields are present. Here’s mine. Hover a field to light its bit.

MTI 0110 · Authorization response Cardholder YASH VERMA
  1. DE 002Primary accountCard payments · gateways · aggregation
  2. DE 003Processing codeEMVCo 3DS · tokenization · smart routing
  3. DE 004AmountPlatforms processing $10B+ a year
  4. DE 007Transmission timeOct 2019 → now · 7+ years
  5. DE 022Entry modeTap · chip · e-commerce · REST APIs (JSON / XML)
  6. DE 039Response code00 · SAFe® 6 POPM certified
  7. DE 041TerminalEMVCo & scheme certification · issuer and acquirer integration
  8. DE 048Additional dataSQL · Power BI · Tableau · QuickSight · Grafana
  9. DE 055ICC dataISO 8583 · ISO 20022 · RBI · EMVCo
  10. DE 060ReservedAI · PMOS · Agent HQ · multi-agent orchestration
  11. DE 063Private usePRDs · roadmaps · UAT · launches · stakeholders

Your turn

Let’s build the next yes.

Hiring for payments, platform or fintech product? I’d love to hear what you’re building.

Email me→ LinkedIn ↗ GitHub ↗

YASH VERMA

PAYMENTS · PRODUCT

PUNE, INDIA

MERCHANTYOUR TEAM

TXN TYPEHIRE

AMOUNT7+ YRS EXP

CARD•••• 2025

AUTH CODEPM2019

STAN000300

RESP00

APPROVED

[email protected]

in/vermayash11

github.com/hellraiser-11

THANK YOU · PLEASE VISIT AGAIN

Approved
Download full résumé (PDF) ↓