Raghuram Kumar

Architected and deployed a secure, enterprise-grade LLMOps platform using Ollama and Qwen2.5-Coder, enabling self-hosted AI-assisted development for 50+ engineers while ensuring complete source-code privacy and regulatory compliance. Implemented ML lifecycle management with MLflow for experiment tracking, model registry, versioning, and automated promotion to production, reducing deployment timelines by over 60%. Built and operated Kubernetes-based ML inference clusters with GPU scheduling, horizontal auto-scaling, and workload isolation for high-performance LLM inference. Embedded DevSecOps controls across ML pipelines, including vulnerability scanning with Trivy, secrets management via HashiCorp Vault, and hardened Kubernetes configurations. Managed hybrid enterprise infrastructure spanning 20+ Windows and Linux servers, SQL databases, and application hosting platforms (Tomcat, IIS), ensuring reliability, backups, and disaster recovery readiness. Administered Azure DevOps Server and

Delivered L3 technical support for enterprise-scale security products including FortiGate firewalls, VPN gateways, IDS/IPS, and SD-WAN solutions for global customers. Investigated and resolved complex network security incidents involving deep packet inspection, SSL/TLS inspection, routing, and high-availability clusters. Developed Python-based automation tools for log parsing, diagnostics, and issue triage, reducing mean time to resolution by approximately 40%. Collaborated with DevOps and platform teams to integrate security validation and compliance checks into CI/CD pipelines using Azure DevOps and shell scripting. Provided guidance on network hardening, secure connectivity, and enterprise firewall design for regulated environments.

Supported day-to-day operations of enterprise networking infrastructure including routers, switches, wireless systems, and VoIP platforms. Diagnosed connectivity, performance, and configuration issues, ensuring minimal service disruption for business customers. Escalated and coordinated complex incidents with network engineering teams, contributing to improved operational response workflows.