I build secure, event-driven backend systems and production-minded AI services with Java.
Based in India · Open to Java Backend, Spring Boot, Microservices, and Distributed Systems roles
| Core strength | Designing reliable Java services, REST APIs, asynchronous workflows, and data-intensive systems |
| Backend | Java 21, Spring Boot, Spring Security, Spring Cloud, Hibernate/JPA, Maven |
| Architecture | Microservices, event-driven systems, API gateways, idempotency, distributed transactions |
| Messaging & data | Apache Kafka, Redis, PostgreSQL, MySQL |
| Cloud & delivery | AWS, Docker, GitHub Actions, Jenkins, Prometheus, Grafana |
| Applied AI | Spring AI, LangChain4j, RAG, MCP, evaluation, guardrails, human approval workflows |
SentinelPay — real-time payment fraud detection
An event-driven payment platform combining transaction authorization, layered fraud detection, observability, and a React operator experience.
- Spring Boot microservices with Kafka, Redis, MySQL, Docker, Prometheus, and Grafana
- Research prototype measured at 95.2% accuracy, 94.2% F1, ~50 ms latency, and 1,200 TPS
- Published research paper, live demo, and v0.1.0 preview release
SentinelOps AI — explainable incident response
A safety-first incident assistant that grounds recommendations in runbooks and operational signals.
- Multi-agent investigation with transparent evidence and human approval gates
- Java, Spring Boot, RAG, Prometheus, and Docker
- Designed to recommend controlled actions instead of making unsafe autonomous changes
ClauseLens AI — evidence-first policy intelligence
A document intelligence service for analysing policies and contracts without losing source traceability.
- PDF ingestion, metadata-scoped retrieval, citations, and compliance-risk detection
- Java, Spring Boot, PDFBox, vector search, and Docker
- Every result preserves evidence for review and audit
EvalForge AI — RAG regression testing
A deterministic evaluation service for detecting retrieval and answer-quality regressions before release.
- Baseline-versus-candidate quality gates with reproducible scoring
- Java, Spring Boot, RAG evaluation, and AI testing
- Treats AI quality as an engineering signal suitable for CI/CD
View all repositories → · View portfolio →
Recruiters and collaborators can verify current build health and implementation evidence directly:
- SentinelPay: CI · releases · runtime demo
- VoyageWeaver ADK: repository · CI
- ShopOps Copilot: repository · CI
- ClauseLens AI: repository · CI
Each repository documents its supported runtime path, tests, container workflow, and deployment assumptions.
- Published research: hybrid AI-assisted fraud detection for digital payment systems
- Open source: merged documentation contributions to Higher-Kinded-J — EitherPath and MaybePath
- Production habits: CI, dependency automation, security policies, contribution guides, release notes, observability, and reproducible local environments
- Problem solving: active LeetCode profile and a validated multi-language solution catalogue
- Reliable, observable Spring Boot microservices
- Event-driven workflows with Kafka and resilient data boundaries
- Secure API design, authentication, authorization, and failure handling
- Production AI integration with evaluation, citations, privacy controls, and human oversight
I am open to backend engineering opportunities where Java, distributed systems, reliability, and thoughtful system design matter.
- Email: shanmugakannan7549@gmail.com
- LinkedIn: linkedin.com/in/k-thillai-nirmal
- Portfolio: thillainirmal-tech.dev
Available for Java Backend Engineer · Spring Boot Developer · Microservices Engineer roles