About Me
Security-focused DevOps / DevSecOps Engineer with 3+ years of hands-on experience designing, automating, and securing cloud-native infrastructure on AWS, Azure, and GCP. Proven track record embedding security into every layer of the platform — IAM least-privilege design, Kubernetes RBAC and network policies, SAST/DAST, container and dependency scanning, secrets management, and SSL/TLS — while delivering Infrastructure as Code (Terraform), CI/CD pipelines, and GitOps-based deployments. Extends security practice into GPU-enabled clusters (NVIDIA) and end-to-end MLOps workflows across SageMaker, Azure Machine Learning, and Google Cloud Vertex AI. Reduced idle compute costs by ~40% through event-driven autoscaling and cut node spend via Spot provisioning. Currently expanding expertise in Apache Spark and data engineering for production-grade platform work.
Skills & Technologies
Work Experience
DevOps / DevSecOps Engineer
Led full lifecycle Kubernetes (EKS and self-hosted) deployments with GitOps and Infrastructure as Code using Terraform. Implemented cost-optimized autoscaling with KEDA and Karpenter reducing idle compute costs ~40%. Managed NVIDIA GPU clusters for ML workloads, deployed ML models using Triton and KServe integrated with major cloud ML platforms. Hardened Kubernetes and cloud infrastructure security with IAM least-privilege, RBAC, namespace isolation, and secrets management. Built CI/CD pipelines with Jenkins, GitHub Actions, AWS CodePipeline incorporating SAST/DAST and container scanning enforcing security gating. Provisioned and optimized AWS infra including VPC, EC2, ALB/NLB, Route53, RDS PostgreSQL, MongoDB Atlas, and S3 with lifecycle policies. Contributed to AI-assisted log analysis using LLMs on EKS and implemented observability with Prometheus, Grafana, and ELK Stack.
Projects
GPU-Accelerated ML Inference Platform on Kubernetes
Architected a multi-tenant Kubernetes cluster with NVIDIA GPU Operator enabling CUDA workloads. Deployed Triton Inference Server for high-throughput model serving (batch + streaming); implemented auto-scaling based on GPU utilisation metrics via Prometheus Adapter and HPA.
End-to-End ETL & MLOps Pipeline
Built an automated data pipeline from S3 ingestion through AWS Glue PySpark transforms to RDS warehouse and feature store. Trained and tracked ML models with MLflow; deployed via KServe on EKS with canary rollout managed by ArgoCD.
LLM-Based Log Intelligence System
Developed a containerised log-analysis service using open-source LLMs (LLaMA-based) served via Triton on GPU nodes; integrated with ElasticSearch for vector-indexed log storage and real-time anomaly alerting.
Zero-Trust DevSecOps Pipeline
Implemented a shift-left security pipeline with SonarQube SAST, Trivy image scanning, OWASP ZAP DAST, and Snyk dependency audits; blocked deployments on critical CVEs and published compliance reports to S3.
Education
Master of Computer Applications
Computer Applications
Bachelor of Computer Applications
Computer Applications