Madhan Kumar R
Madhan Kumar R

Madhan Kumar R

Senior Backend Engineer, based in Coimbatore, India

I build the backend systems that move money in real time.

  • Node.js
  • Apache Kafka
  • Microservices
  • MongoDB
  • Card payments
  • Tokenization
  • UPI
  • Finacle core banking

6.5+ years in banking, payments and healthcare

Download resumeLinkedInGitHub
integrations shippedbanking and payments
  • Card schemesMastercard, Visa, RuPay, Jaywan
  • PaymentsUPI, PhonePe, Cashfree, net banking
  • Core bankingFinacle
Live in production at a major UAE bank and a payment gateway

6.5+

years building backend systems

6

engineers led on a payment gateway

2,000+

concurrent transactions sustained

100,000+

users served on a healthcare platform

Experience

  1. Dec 2024 – Present

    Senior Backend Engineer at D4Insight

    Client: Emirates NBD

    • Process Finacle events on Kafka for debit card issuance, delivery and lifecycle updates.
    • Rebuilt the card delivery flow as event-driven Kafka microservices.
    • Moved Finacle credit card bill payment sync from a next-day (T+1) batch to instant updates.
    • Made these flows fault-tolerant with 3+ automatic retries per event and a cron-based retry job.
    • Launched wallet-token management across Mastercard, Visa and Jaywan (UAE domestic scheme).
    • Built secure REST APIs with audit trails, aligned with banking compliance and SLA requirements.
    • Building the Mastercard rewards claim integration.

  2. Mar 2022 – Nov 2024

    Full Stack Engineer at Techno Consulting Services

    Client: CredoPay (payment gateway)

    • Led a team of 6 engineers; code reviews and automated testing cut production bugs by 40%.
    • Integrated Mastercard, Visa, RuPay (NPCI), PhonePe and Cashfree into a multi-network gateway on NestJS.
    • Built UPI Collect, UPI Intent, card and net-banking flows with AES/RSA encryption and tokenization.
    • Designed Kafka streaming with Redis and Bull queues, sustaining 2,000+ concurrent transactions.

  3. May 2020 – Feb 2022

    Software Engineer at Techno Consulting Services

    Client: Healthcare SaaS platform

    • Built FastAPI and PostgreSQL services for a platform serving 100,000+ healthcare professionals.
    • Developed a candidate-to-job matching algorithm that cut screening time by 35%.
    • Designed a Dockerized CI/CD pipeline that halved deployment time.

How I design event-driven systems

The pattern I use for systems that can't lose an update. Failures are expected, so ordering, duplicates, recovery and checking are designed in from the start.

Producer

An upstream system emits an event

Each event carries a unique ID

Message broker

Kafka keeps events durable and replayable

Partition key per entity keeps its events in order

Consumer service

A Node.js service processes the event

Idempotency check by event ID, so a retry never applies twice

Downstream system

The update is available right away

after 3+ retries with backoff

Scheduled reprocessor

Re-publishes parked events to the broker once the cause is fixed

Dead-letter queue

Events that still fail are parked, not lost

Reconciliation job

Compares records against the source of truth on a schedule

Mismatches are flagged before customers notice

Observability

Grafana dashboards and ELK logs

Alerts on consumer lag and dead-letter queue size

Producer

An upstream system emits an event

Each event carries a unique ID

Message broker

Kafka keeps events durable and replayable

Partition key per entity keeps its events in order

Consumer service

A Node.js service processes the event

Idempotency check by event ID, so a retry never applies twice

Downstream system

The update is available right away

Case studies

Write-ups of how I designed two systems, and the trade-offs.

Card delivery tracking

From next-day batch updates to real-time events on Kafka, with retries and a cron safety net.

Node.js · Kafka · MongoDB

Read the case study

A multi-network payment gateway

Mastercard, Visa, RuPay, PhonePe, Cashfree and UPI behind one gateway, built by a team of 6.

NestJS · Kafka · Redis

Write-up coming soon

Projects

Vasool Money

Live

A loan-tracking app for small lending businesses. Owners, staff and collection agents manage customers, loans and daily collections across several organizations, with reports for expected versus collected amounts.

  • Role-based access for owners, staff and agents
  • Agents record collections offline and sync later
  • Android and web, in English and Tamil
Visit vasoolmoney.com

Skills

Languages and frameworks
Node.js, TypeScript, NestJS, Express, Python, FastAPI, Vue.js
Messaging
Apache Kafka, Bull queues, event-driven microservices
Data
MongoDB, PostgreSQL, Redis
Payments and security
Card schemes, tokenization, UPI, Finacle, AES/RSA, JWT
Delivery and operations
Jenkins, GitHub Actions, Docker, AWS, Grafana, ELK Stack, Playwright

Get in touch

Open to senior backend roles at fintech product companies, remote or hybrid.

Download resumeLinkedInGitHub