KJ
Currently at Coursera · Open to interesting problems

Kajal Jain
Backend engineer, built for scale

I design distributed backends — async pipelines, gateways, isolation, on-call. Built to survive IPL-scale traffic and ship revenue at PhonePe, Dream11 and Coursera.

Download ResumeContact me
India · Remote-friendly
5+years shipping backend
250M+users served
20M+concurrent traffic handled
About

Backend, at the shape of the problem.

I'm a backend engineer with 5+ years of experience designing distributed systems across fintech, gaming, sports-tech and edtech — most recently at Coursera, Dream11, PhonePe and PlaySimple Games.

My focus is on the unglamorous parts of scale: async pipelines, event-driven architectures, traffic isolation, on-call reliability and the API contracts that let dozens of teams ship independently. I care about the boring guarantees — idempotency, back-pressure, graceful degradation — because they're what let a system survive an IPL-scale traffic spike or a Tier-1 gateway migration.

I like sitting close to the product. The best backend work I've shipped moved a revenue number, unlocked a new integration surface, or removed a class of on-call pages entirely.

Distributed Systems
Event-driven, multi-tenant, high-throughput
Backend at Scale
250M+ users, 20M+ concurrent traffic
System Design
Async pipelines, gateways, isolation
Cloud & Data
AWS, Kafka, Redis, MySQL, ClickHouse
The journey

Scale, ownership, and the hard parts of backend.

Each role — a deliberate step up in scale and ownership.

C
Coursera
Software Engineer II
Mar 2026 — PresentRemote
JavaKafkaSQSRedisgRPCAirflow
What I owned

Building the enterprise integrations platform syncing Coursera with 30+ external LMS ecosystems.

Signature wins
30+ LMS integrations

Designed and shipped a scalable content sync pipeline powering integrations with external learning platforms.

100% drops eliminated

Diagnosed and resolved a long-standing curriculum sync failure, eliminating consumer drops and restoring cross-org update propagation.

15+ enterprise platforms

Architected async backend services on Kafka, SQS, Redis and gRPC, unlocking new integration surfaces for enterprise customers.

Also on my plate

Built Airflow-based reverse-sync automation for Degreed Required Learning, removing manual ops across eligible organizations.

Led on-call rotations and incident response; drove RCAs and shipped automation that measurably reduced MTTR.

Featured Work

Systems, not features.

Four backend problems worth talking about.

Federated GraphQL Gateway

Self-serve deploys for a Tier-1 API gateway

Problem
Subgraph teams were blocked behind a single monolithic gateway deploy — routine changes took 3–4 day release trains.
Architecture
Migrated to Apollo-style federation with independently deployable subgraphs and schema-composition checks in CI.
GraphQL FederationJavaAWSCI/CD
3–4 days → self-serve
Release cycle
10+
Subgraph teams

Multi-Tenant Comms Platform

One platform. Five channels. Ten+ teams.

Problem
Every product team at Dream Sports was building its own notification stack across five channels.
Architecture
Multi-tenant messaging platform with per-tenant rate limits, retries, provider abstraction and push batching on CleverTap.
JavaKafkaRedisCleverTapAWS
5
Channels unified
Eliminated
Manual campaign ops

LMS Content Sync Pipeline

Reliable sync across 30+ external learning platforms

Problem
Enterprise customers needed Coursera content mirrored bidirectionally with their own LMS — with reliability and ordering guarantees.
Architecture
Async pipeline over Kafka + SQS with Redis-backed state, gRPC service boundaries, and Airflow DAGs for reverse-sync.
JavaKafkaSQSRedisgRPCAirflow
30+
LMS integrations
0
Consumer drops

Recharges Traffic Isolation

Surviving IPL and World Cup on the same infra

Problem
A single noisy operator could saturate shared thread pools and cascade failures across the entire BU.
Architecture
Per-operator custom thread pools and queue partitioning, with circuit breakers and priority lanes for revenue-critical flows.
JavaSpring BootAerospikeKafka
IPL / World Cup
Peak event
Per-operator
Blast radius
Toolkit

The stack I reach for.

Things I've shipped to production and would happily debug at 3am.

Backend & Architecture
  • Distributed Systems
  • Microservices
  • Event-Driven Architecture
  • Real-Time Systems
  • System Design
  • API Gateways
Languages & Frameworks
  • Java
  • C++
  • Python
  • SQL
  • Spring Boot
  • Dropwizard
  • GraphQL
Messaging & Streaming
  • Kafka
  • AWS SQS
  • RabbitMQ
Databases & Storage
  • MySQL
  • Aurora
  • MariaDB
  • ClickHouse
  • Aerospike
  • Redis
  • Elasticsearch
  • AWS S3
Cloud & Infra
  • AWS
  • Airflow
  • gRPC
  • CI/CD
AI / LLM Systems
  • LLM API Integration
  • MCP Servers
Impact

Where the numbers moved.

Selected outcomes worth remembering — from revenue lifts to reliability wins.

20M+ concurrent

IPL-scale on-call ownership

Primary on-call for backend services during India's largest live sports traffic events — live debugging, mitigation, and post-incident remediation.

250M+ users

Engagement infra for 250M+ users

Kept the systems powering the Dream11 user base reliable through peak traffic and complex multi-tenant workloads.

3–4d → self-serve

Federated gateway migration

Led adoption of federated GraphQL across 10+ subgraph teams, moving a Tier-1 API gateway from centralized to self-serve deploys.

+12% recharges · +7% DTH

Revenue-moving integrations

Solo-led Vodafone direct integration at PhonePe and orchestrated Tata Sky DTH direct integration.

+161% Jio · +61% Airtel · +41% VI

New-customer growth on prepaid→postpaid

Shipped a redirection flow across all operators that unlocked a step-change in new customer acquisition.

42% faster

Cold-start rebuild on a 10M+ user game

Reworked asset bundling on Word Wars, cutting app load time from 14s to 8s on live devices.

Education

Foundations.

B.Tech, Computer Science

Lovely Professional University · Jul 2017 — Jul 2021

CGPA
9.30 / 10
Contact

Let's build something worth shipping.

I'm always up to talk about distributed systems, on-call horror stories, or a role where backend actually matters.

Emailkajaljain0509@gmail.comLinkedInkajal-jain

Prefer email? Drop me a note at kajaljain0509@gmail.com. I read every message and usually reply within a day.

Currently based in India · Remote-friendly.