Looking for a job

Harsh Kankariya

@harshkankariya

Built scalable backend systems and REST APIs with Java, Spring Boot, Node.js and AWS, with hands-on experience in RAG and Agentic AI solutions.

India

Message

What I'm looking for

I'm looking for Java/Spring Boot backend and microservices roles where I can build scalable, production-grade systems and continue growing in agentic AI, Kubernetes, Terraform, and advanced LLM orchestration.

At Avalara, I architected Spring Boot microservices and curation pipelines for taxonomy processing across 10+ countries. I processed 1–5 GB AWS S3 files, used PostgreSQL as the source of truth, and supported mandates, code lists, and specifications with 90%+ SonarQube coverage.

I improved production reliability with Redis token caching, Docker deployments on AWS ECS, Prometheus monitoring, and GitLab CI/CD quality gates. As an on-call engineer, I resolved 15+ production issues each month and helped cut MTTR by 20%; I also reduced batch runtime from 60 to 10 minutes using Node.js Worker Threads and BullMQ.

I'm currently building a Field Survey & Grievance Portal with Spring Boot, PostgreSQL, role-based JWT authentication, QR-based citizen login, and REST APIs. I also build RAG and LLM applications, including a GitHub repository assistant using Python, FastAPI, LangChain, FAISS, and the OpenAI API.

Experience

Work history, roles, and key accomplishments

Current

Software Engineer (Contract)

Current

Freelance

Apr 2026 - Present (5 months)

Building a full-stack Field Survey & Grievance Portal with Spring Boot and PostgreSQL, implementing role-based JWT auth and QR-code-based citizen login. Designed RESTful APIs for survey creation, assignment, and status tracking.

Spring Boot PostgreSQL JWT REST APIs Fullstack Development QR Code Integration

Software Development Engineer

Avalara

Jan 2025 - Mar 2026 (1 year 2 months)

Architected a Content-Consumer microservice processing 1-5 GB taxonomy files from AWS S3, supporting 10+ countries with zero core-code changes. Implemented Redis token caching achieving 95% cache hit rate and reduced batch job runtime from 60 to 10 minutes using Node.js Worker Threads and BullMQ.