Rajeev Kushwaha — Senior Software Engineer · Java · Kafka

01 · Selected systems

Built for the moments
when failure isn't an option.

Payments at peak load. Healthcare claims that must reconcile to the penny. Retail orders that can't arrive out of sequence. Three systems that shaped how I engineer — each one measured by a number, not an adjective.

Financial services · Synchrony

Real-time transaction processing at peak load

Event-driven microservices processing millions of daily financial transactions. Engineered Kafka consumer backpressure with pause–resume strategies so services degrade gracefully under peak volume — instead of paging someone at 2 a.m.

▸ Batch pipeline re-platformed to Lambda + S3: 35 min → 6 min

spring-boot · kafka-streams · avro · aws-lambda · redis · oracle

Healthcare · DXC Technology

Claims reconciliation across millions of records

Enterprise reconciliation engine comparing Kafka streams against Oracle at scale to guarantee data accuracy — paired with an Angular ops portal that lets operations inspect DLQ messages and trigger automated retries themselves.

▸ Query tuning via explain plans & indexing: minutes → seconds

spring-batch · kafka · oracle-plsql · angular · aws-lambda · k8s

Retail · Target

Ordered events & self-healing failure paths

Kafka partitioning strategies that preserved strict transaction ordering across distributed services, plus a custom Dead Letter Queue framework that isolates and recovers failed messages automatically — no manual triage.

▸ Failed-message recovery: manual triage → automated

kafka · spring-boot · ngrx · react · dynamodb · redis

02 · How I engineer

Senior means the system
is your problem — all of it.

Anyone can write code that works on a good day. The job is designing for the bad ones — and owning what happens next.

⚡

Design for failure first

Backpressure handling, Dead Letter Queues, idempotent consumers, automated retries — I design the failure path before the happy path. In systems moving money and medical claims, graceful degradation isn't a feature; it's the contract.

📐

Measure, don't guess

Explain plans before rewrites. Prometheus and Grafana before "it feels slow." Every optimization I ship has a number attached — 35 minutes to 6, minutes to seconds — because performance claims without baselines are just opinions.

🎯

Own it end to end

Requirements, design, build, deploy, monitor, support. At Synchrony, DXC, and Target I carried services from whiteboard to production pager — working directly with architects, product owners, QA, and ops. No handoffs, no "not my layer."

03 · Experience

Nine years. Four industries.
One standard: production-grade.

Synchrony Financial

Senior Software Engineer

SEP 2023 — PRESENT

Charlotte, NC

Real-time financial transaction processing platform

Own end-to-end delivery of high-volume transaction services — requirements through design, build, deployment, and production support.
Designed event-driven microservices with Spring Boot + Kafka processing millions of financial transactions in real time.
Architected consumer backpressure handling (pause–resume) that prevented failures during peak transaction loads.
Migrated legacy batch systems to AWS Lambda + S3, cutting processing time from 35 to 6 minutes.
Built secure messaging with encryption, masking, and audit logging; eliminated hardcoded credentials via Secrets Manager.
Shipped observability with Prometheus, Grafana, ELK, and Splunk; automated CI/CD on Jenkins, Docker, ECS, Kubernetes.

DXC Technology

Senior Software Engineer

NOV 2021 — AUG 2023

Plano, TX

Healthcare claims processing & data reconciliation platform

Led end-to-end ownership of claims processing services from design through production, across business, QA, DBA, and cloud teams.
Built reconciliation tools comparing millions of records between Kafka streams and Oracle to guarantee data accuracy.
Developed Angular support portals for DLQ inspection and automated message retries.
Cut query execution from minutes to seconds through explain-plan analysis, indexing strategy, and PL/SQL tuning.
Implemented AWS Lambda validation and extraction workflows; containerized deployments on Kubernetes and Docker.

Target Corporation

Software Engineer

DEC 2018 — OCT 2021

Minneapolis, MN

Retail transaction & order management platform

Owned retail order and transaction services end-to-end, from analysis and design through deployment and production support.
Designed Kafka event architectures with strict partitioning to preserve transaction ordering and consistency.
Built a custom Dead Letter Queue framework for automated isolation and recovery of failed messages.
Delivered Angular + NgRx enterprise frontends and React dashboards for real-time transaction analytics.
Shipped serverless AWS Lambda workflows for automated financial processing and notifications.

Morningstar

Software Engineer

MAY 2017 — NOV 2018

Chicago, IL

Real-time market data processing platform

Built real-time ingestion services for high-frequency market data feeds.
Implemented multithreaded processing models to raise throughput and cut latency.
Designed resilient recovery mechanisms validated with JUnit/Mockito fault-injection testing.
Managed AWS infrastructure: API Gateway, S3, IAM, CloudWatch, SNS, SQS, SES; archival on S3 Glacier and EFS.

I build systems that
move millions —
in real time.

Built for the moments
when failure isn't an option.

Real-time transaction processing at peak load

Claims reconciliation across millions of records

Ordered events & self-healing failure paths

Senior means the system
is your problem — all of it.

Design for failure first

Measure, don't guess

Own it end to end

Nine years. Four industries.
One standard: production-grade.

The stack behind the systems.

Certified where it counts.

Built for the momentswhen failure isn't an option.

Real-time transaction processing at peak load

Claims reconciliation across millions of records

Ordered events & self-healing failure paths

Senior means the systemis your problem — all of it.

Design for failure first

Measure, don't guess

Own it end to end

Nine years. Four industries.One standard: production-grade.

The stack behind the systems.

Certified where it counts.

Built for the moments
when failure isn't an option.

Senior means the system
is your problem — all of it.

Nine years. Four industries.
One standard: production-grade.