Currently shipping Toronto, ON

Sreyas
Kurup.

Payments, agents, and the rails money will quietly run on.

I work on enterprise payments at IBM. Days I'm building for a leading U.S. card network and a Fortune 500 insurer, with side work pulling in central banks and large American institutions. Nights I'm writing about agentic payments, stablecoin settlement, and figuring out what comes after SWIFT.

Scroll

I make payments work for the people moving billions through them.

Most payments products fail in the unglamorous middle: between a bank's mainframe and a treasury team's morning coffee, in the gap between what an exec promised in a deck and what compliance will actually sign off on.

That gap is where I live. I take the messy operator reality of treasurers, ops leads, integrators, and partner banks, and I turn it into product that actually ships.

  • BasedToronto
  • Day JobPayments Transformation Consultant, IBM
  • EngagementsLeading U.S. card network, Fortune 500 insurer
  • FocusAgentic payments, stablecoins, ISO 20022, SWIFT
  • SchooledMath, University of Waterloo
  • On the sideEdTech, writing, a couple of bets I'll tell you about below
01 / Track Record

The work.

Five years across consulting, payments product, and research. Click any row for the long version.

2023 → Now

Payments Transformation Consultant

IBM Consulting, Toronto
GTM, delivery, partnerships
+
Mandate

I run go-to-market and delivery across IBM's payments platforms: real-time rails, wires, EFT, cheques, and the API layer that ties them together. The job spans the full arc: partnerships, conference circuits, RFPs, client onboarding, contract negotiation, defining MVPs, and squeezing the P&L.

The piece I'm proudest of right now is agentic payments and AP2 protocol adoption. I helped define the use cases, wrote the requirements, and pulled in IBM Research to stand up proof-of-concepts with our partner ecosystem. Most of the industry is still pattern-matching agentic to chatbots. We're treating it as a new payments rail with its own auth model, reconciliation surface, and risk profile.

I also run our DLT, stablecoin, and CBDC work for banks, and design asset-based consulting plays that turn one-off engagements into PaaS and SaaS revenue lines.

AP2Agentic AIDLTStablecoinsCBDCRTP / FedNowSalesforcePartnerships
2024 → Now

Lead Payments Consultant

Leading U.S. card network (remote)
SWIFT, ISO 20022, treasury
+
Mandate

Trusted advisor to Directors, VPs and senior executives. I designed and implemented the SWIFT integration for near real-time liquidity reporting, giving treasury proactive visibility across $1T+ in annual global cashflows.

Owned the MT to MX migration for liquidity reporting under ISO 20022: onboarded partner banks, enriched data granularity, built reconciliation logic, and tied the whole thing back to Basel III LCR/NSFR compliance so the regulators stay quiet.

I run the global banking integration side as well, defining technical requirements, managing SWIFT setups, and untangling integration issues across 10+ countries of banking partners.

SWIFTISO 20022 MXLiquidity ReportingBasel IIIReconciliationGlobal Treasury
2025 → Now

Product Manager, Payments Platform

Fortune 500 U.S. insurer (remote)
Post M&A platform, 0 to 1
+
Mandate

Defined a 0 to 1 enterprise payments platform after a major acquisition. The job: unify ACH and check disbursements across two acquired benefits systems, decouple legacy apps, and stitch front and back-office together with an API orchestration layer.

Implemented ACH flows (CCD, CTX, PPD) with automated handling for returns and NOCs. Stood up exception queues and remediation playbooks that killed off the spreadsheets ops was using to triage failures.

Ran discovery workshops, translated findings into epics, user stories, and acceptance criteria, and kept leadership, Treasury, GL, and IT Ops aligned on scope and sequencing. Compliance work covers tokenization, fraud checks, OFAC screening, and scheme rules baked into the release audit gates.

ACHAPI OrchestrationPost M&A IntegrationOFACTokenizationAgileDiscovery
2021 → 2023

Research Supervisor

Statistical Consulting and Research Unit, University of Waterloo
Survey research, team lead
+
Mandate

Ran a research team for the International Tobacco Control project, the largest smokers' survey research program in Canada. Findings fed legislative changes on what tobacco retailers and manufacturers could put on a shelf or a label.

Cleaned and harmonized multi-wave datasets, built respondent-behavior segmentations, and ran latitudinal and longitudinal trend analyses that ended up in policy briefs. On the people side: training, incident reports, performance reviews, the team-building events nobody asked for but everyone needed.

Survey ResearchRSQLPolicyTeam Lead
02 / What's Next

Three bets on the future of payments.

Working titles, working prototypes, working hypotheses. None of these are finished. All of them should exist.

01
Codename / PULSE

Agent to agent settlement, at sub-second finality.

A payment layer purpose-built for autonomous agents. Every API call between AI services becomes a metered, atomic financial transaction with cryptographic intent, replayable receipts, and policy-aware routing. The thesis: when agents start transacting with each other in volume, the existing card networks fall over within a quarter. Pulse is the rail that doesn't.
Stage: prototype, partner discovery
02
Codename / LATTICE

The router that picks the rail for you.

A real-time payments router that decides whether each transaction goes ACH, RTP, FedNow, wire, card, or stablecoin, based on live cost, risk score, counterparty SLA, and regulatory posture. Treasurers stop choosing rails. Lattice does it per payment, per microsecond, with a full audit trail. Imagine BGP for money.
Stage: architecture, design partner conversations
03
Codename / VAULT ZERO

Privacy-preserving B2B settlement using ZK proofs.

Cross-border B2B payments that satisfy KYB, sanctions screening, and travel-rule reporting without exposing counterparty PII to every node in the chain. ZK proofs for compliance attestations, attested intent for routing, on-chain settlement for finality. Built for the regulator who hates surprises and the CFO who hates leaks.
Stage: whitepaper draft, regulator conversations
03 / Thought Dump

Things I wrote, mostly opinions.

Conference notes, hot takes, late-night drafts. Some have aged well. Some haven't.

Field NotesNACHA SFP 2026
The panels at NACHA were boilerplate AI, stablecoins, ISO and real-time, but the credit union reps in the hallways had the actual story.

Largest attendee group at the whole conference, and almost every one of them told me the same thing: completely dependent on the big banks they connect through, the service is mediocre, and the processes are still humans pushing manual forms back and forth in 2026. Someone needs to platformize the credit union to bank pipe and retire the form-pushers entirely. That's the company I'd back tomorrow.

Field NotesPayments Canada 2026
The biggest booth at Payments Canada 2026 belonged to a company nobody has heard of called dobusiness.com.

Not a bank. Not a Schedule I incumbent. Not even one of the usual fintechs you'd expect to dominate a Canadian payments floor. The center of gravity is quietly shifting to the PSPs, and the banks haven't noticed yet. By the time they do, the merchant relationship will already belong to someone else.

Hot TakeAgentic
Agentic payments will eat the card networks before 2030, and the networks won't rebuild in time.

Cards optimized for a world where humans tap once and trust the merchant. Agents will transact thousands of times per minute, on behalf of someone who isn't watching, with intent that has to be machine-verifiable. The networks aren't built for that. They'll either rebuild or be routed around. They won't rebuild.

Hot TakeStablecoins
Stablecoins won't replace SWIFT, they'll quietly absorb it from the inside.

Everyone keeps framing this as a fight. It isn't. SWIFT is messaging. Stablecoins are settlement. The future is MX messages carrying on-chain transaction IDs, with SWIFT operators acting as the trust layer over a stablecoin rail. The banks that figure this out first will look like they invented cross-border. They didn't. They just stopped resisting it.

Protocol WatchAP2
AP2 is the most underrated protocol of the decade, and most payments people haven't read the spec.

Auth for agentic commerce, with intent signed by a human, delegated to an agent, and verifiable by a merchant. It's the missing link between "AI can browse for me" and "AI can actually buy for me without breaking dispute frameworks." Most people in payments haven't read the spec. They will, in about six months, in a panic.

SpicyCrypto
Crypto's only real product is settlement, and everything else is theatre.

Strip out the NFTs, the memecoins, the discourse. What's left is an extremely good settlement layer that doesn't sleep on weekends. That alone is worth a trillion-dollar market. The industry's biggest enemy isn't regulators. It's its own marketing team.

04 / On the Side

Freelance & side bets.

Things I built outside the day job, usually because nobody else was building them.

Freelance / EdTech Client

A virtual internship simulation platform.

An EdTech client came to me with a simple problem: the next generation of students wanted internships but couldn't get them, and the companies that mattered weren't going to start hiring 16 year olds. I helped them design and build a virtual internship simulation platform that compresses real-world early-career work into structured, scored modules.

The simulations draw from my own internship experiences across consulting, payments, and research: real artifacts, real ambiguity, real decisions, with feedback loops built in. I owned product thinking on the simulation engine, the rubric scoring layer, and the storyline arc that keeps a 17 year old engaged past day three. Now used by students across multiple countries and rolled into school curriculums.

Product StrategySimulation DesignEdTechCurriculumUX Writing
Writing

Stablecoin Settlement via SWIFT and Circle's on-chain network.

A framework for plugging SWIFT's messaging fabric directly into Circle's on-chain settlement network. The premise: real-time cross-border settlement without correspondent banks, by mapping on-chain transaction identifiers to MX messages so SWIFT operators and web3 wallets speak the same language. Published on Medium, read by more banks than I expected.

SWIFTCircle CPNStablecoinsISO 20022Cross Border
05 / Get in Touch

Let's talk.

If you're working on payments, agents, stablecoins, or anything that sits next to them: drop a note. I read everything.

LocationToronto, ON

Best for: partnerships, payments product collaborations, speaking, deep-dive coffee on agentic settlement. Not for: anything that starts with "quick AI question."

Sent. I'll get back within a couple of days.