Building production-ready GenAI/LLM solutions and scalable cloud infrastructure on AWS & Azure. Automating end-to-end pipelines for AI engineering.
I am a Cloud & MLOps Engineer based in Hof, Germany, with cross-cloud experience in AWS and Azure. I architect and deploy scalable software systems, containerised services, and GPU-accelerated ML workloads on Kubernetes (EKS & AKS).
Experienced in building production-ready GenAI/LLM solutions, MCP servers, and automated CI/CD pipelines with comprehensive DevSecOps controls.
Oneture Technologies · Mumbai, India · 09/2024 – 09/2026
AWS, Microsoft Azure
Docker, Kubernetes (EKS & AKS), Helm, Karpenter, Docker Compose
Azure DevOps, GitLab CI/CD, Jenkins, Terraform, AWS CodePipeline, CloudFormation
FastAPI Model Serving, GPU Inference on EKS, FastMCP, Containerized LLM Deployment
SonarQube, Trivy, OWASP ZAP, Veracode, Twistlock, OWASP Dependency-Check
Prometheus, Grafana, CloudWatch, Splunk
Python, Bash, Git, PostgreSQL, MySQL, Linux, Nginx
Built production-ready GenAI solutions from scratch, including Python-based document conversion services (PDF → Markdown) and reusable MCP servers exposing AI/LLM tools via standardized API calls.
Deployed AI MCP servers and backend services to Azure Kubernetes Service with high availability. Automated multi-repo pipelines with Veracode, SonarQube, Twistlock, and Docker scanning gates.
Improved AWS security posture from ~20% to >85% using native controls, account-level guardrails, and centralized monitoring. Migrated workloads to private subnets with least-privilege access.
Reduced ML inference costs by ~80% by migrating from AWS SageMaker to containerized model serving on EKS with autoscaled GPU nodes. Managed cluster reliability via Prometheus and Grafana.
Thoughts on LLMOps, Agentic AI, and Engineering
Interested in discussing LLMOps, AI Engineering, or have a project in mind? Feel free to reach out.