Istanbul, Türkiye
• Consulted on enterprise-level software modernization, overseeing architecture design and Kubernetes migration. • Developed AI workload pipelines across AWS, GCP, and on-premise GPU stacks, enhancing customer development stacks. Working on self hosted LLMs and managed and fixed performance issues. Enhanced Nvidia GPU performance and developed e2e monitoring systems with BPF. • Designed complex network and virtualization architectures, including site-to-site VPN, encrypted traffic on virtually created networks in pure Linux systems. • Mentored team members during technical incidents, fostering a collaborative and efficient work environment. • Designed hybrid cloud infrastructure with AWS and Azure. Designed zero downtime switch mechanism in any kind of failure and regional failures. • Developed AI based SRE incident first contact response systems and enriched them with AI agents.
Managed database clusters and underlying infrastructure: Administered client database clusters and servers, ensuring high availability and performance with cross teams Optimized system performance: Conducted monitoring and performance tuning on high I/O Linux servers to reduce latency and enhance efficiency. Infrastructure as Code (IaC) and CI/CD management: Maintained and improved internal Infrastructure as Code (IaC) and CI/CD pipelines to streamline deployments. Developed automation tools and APIs: Built and maintained automation tools and production management APIs using Golang, Shell, and Python. Enhanced monitoring and alerting systems: Improved observability by refining monitoring and alerting systems, and developed custom Prometheus exporters in Golang. Implemented cost optimization and security automation: Developed automation and detection systems to improve cost efficiency and security compliance SOC-2.
Manage customers monitoring and incident management system Working on auto scalable systems and multi tenant systems under high traffic on AWS Implement disaster recovery and applied Chaos Scenarios GPU based system tuning