Karunam Goyal
SENIOR BACKEND ENGINEER · DISTRIBUTED SYSTEMS
Senior Backend Engineer with 4.5+ years scaling mission-critical payment systems to 10 million users at CRED and Flipkart. I specialise in designing low-latency, cost-efficient microservices and platforms while integrating AI frameworks to accelerate development cycles.
Currently at CREDbuilding the Metals platform (Gold & Silver) and a new Resident app. Previously shipped Fixed Deposit, CRED Gold rewards, Kuvera mutual fund checkout, and CRED's Broking OMS.
Before CRED, spent 3.5 years at Flipkart — starting as an intern building gRPC migration tooling, then joining full-time on the UPI team. From SDE to SDE2, scaled UPI from zero to 10 million users in 6 months, built the core microservices stack, instant refund infrastructure, and a high-throughput distributed streaming platform.
The lines between backend, infra, and AI are blurring fast. Interested in how distributed systems thinking applies to reliable AI-agent infrastructure — and keeping an early eye on quantum computing as the next paradigm.
- ›Currently building Metals platform (Gold & Silver) and a new Resident app — yet to launch.
- ›Led architecture and delivery of Fixed Deposit platform from scratch — full partner integrations, legal & compliance coordination, successful GTM launch.
- ›Developed Gold Rewards allocation system for CRED Gold onboarding — allocated Rs. 3 crore worth of rewards in the first week, boosting retention KPIs.
- ›Built Checkout Service in Golang, integrating Kuvera, BSE, and CREDPay for a high-throughput, low-latency mutual fund and broking checkout experience.
- ›Architected High-Performance Order Management System (OMS) for CRED's Broking Platform — integrated with exchange test LL setup, ensuring exchange-grade reliability.
- ›Integrated Prometheus monitoring into Account Aggregator, replacing expensive Databricks queries — real-time alerts, improved observability, significant cost savings.
- ›Part of the core Flipkart UPI senior team — achieved 1 million customers in week 1 and 10 million users within 6 months of product launch.
- ›Crafted multiple key microservices using Dropwizard from scratch: User Service, Gateway, UI Adaptation layer, Rewards Manager, Notification service, Auto-pay, Rate-limiter, and Timeout Module — serving 10M customers at high scale.
- ›Implemented a high-performance distributed platform using Apache Storm with Apache Kafka and Pulsar as stream sources for critical async capabilities — 8 onboarded use-cases including UPI live transaction flow.
- ›Conducted first red-line load testing for the refunds system — handled 10x BAU load, identified bottlenecks, tuned for peak sale hours.
- ›Worked on instant refund initiative with Visa Direct, MasterCard Send, and UPI Refunds — reduced SLA from 3-5 business days to 24 hours.
- ›Proactive participation during external partner outages prevented SLA breaches on 30-40k refunds at a time.
- ›Managed critical migrations: VM to Kubernetes, DB version upgrade, DC migration — resulted in 50% fewer on-call incidents.
- ›Part of critical RBI tokenisation compliance program for refunds.
- ›Set up Azkaban Workflow Engine for automating and scheduling multiple jobs.
- ›Worked on a standalone project for future gRPC migrations — automatic conversion of developed REST POJO models to proto3 IDL.
- ›Built a library for creating log summaries and testing conversions when moving existing classes to proto definitions.
Reading · Watching Movies · Cooking · Exploring Different Places
Want to work together or just say hello?