Looking for a job

Rohit Raj Kaushik

@rohitrajkaushik

Message

Backend Software Engineer building scalable distributed microservices and GenAI RAG pipelines.

India

Message

What I'm looking for

I’m looking for a backend role where I can own scalable, production-grade event-driven microservices and build GenAI/RAG pipelines that perform reliably at high concurrency in cloud-native environments.

I’m a Backend Software Engineer with nearly 3 years of experience specializing in distributed systems and high-concurrency microservices. I focus on building scalable, production-ready backend systems with high code quality and reliability, from core API development to event-driven orchestration.

At Carelon Global Solutions, I built GenAI/RAG backend APIs that query OCR-processed insurance claim documents, using chunking, embeddings, semantic retrieval with Qdrant, and structured JSON outputs stored in MongoDB. I also led migrations from legacy orchestration to event-driven ingestion using Redis and RabbitMQ, integrated Apache Kafka for audit-processing workflows, optimized async pipelines to reduce ingestion latency by 40%, and improved API performance by 60% through non-blocking design patterns.

Experience

Work history, roles, and key accomplishments

Current

Software Engineer III

Current

Carelon Global Solutions

Oct 2025 - Present (8 months)

Built GenAI/RAG backend APIs to query OCR-processed insurance claim documents using chunking, embeddings, and semantic retrieval via Qdrant with structured JSON outputs stored in MongoDB. Integrated Apache Kafka for audit-processing event streams, implementing manual offset management in consumers to ensure controlled acknowledgments and prevent message loss during failures.

Python fastAPI Qdrant MongoDB Retrieval Augmented Generation (RAG)Vector Search Apache Kafka RabbitMQ Docker Kubernetes

Associate Software Engineer III

Carelon Global Solutions

Jul 2023 - Oct 2025 (2 years 3 months)

Spearheaded migration from a legacy vendor orchestrator to a custom event-driven ingestion pipeline using Redis and RabbitMQ, increasing concurrent processing capacity 3x. Improved ingestion latency by 40% using the Claim Check Pattern and raised API performance by 60% by removing blocking calls and enabling asynchronous write-behind processing.