Site Reliability Engineer · Linux · Kubernetes · Distributed Systems · Exploring AI Infrastructure

I'm a Site Reliability Engineer who likes knowing why a system behaves the way it does, not just how to operate it. A pod stuck in Pending, a slow request, a flapping node: I want to trace each one down through the container runtime, the kernel, the network and the distributed system around it.

My strongest interests are Operating Systems, Computer Networking and Distributed Systems. Linux and Kubernetes are where I get to study all three in practice.

I build and operate reliable infrastructure today, while exploring how those principles apply to AI-heavy workloads tomorrow.

Direction: SRE → Platform Engineering → AI Infrastructure / AI InfraOps

• Running and improving reliable, observable production systems as an SRE
• Going deeper into Linux internals and Kubernetes (how it works, not only how to use it)
• Automating repetitive operational work with Bash and Ansible
• Building stronger observability practice with Dynatrace and Splunk
• Learning how infrastructure for AI workloads differs from traditional services

Operating Systems & Systems

Cloud & Infrastructure Automation

Containers & Orchestration

DevOps & CI/CD

Observability

Fundamentals I care most about

Operating Systems · Computer Networking · Distributed Systems

AI InfraOps is the practice of designing, operating, automating, scaling, monitoring and securing reliable infrastructure for AI/ML workloads and AI-heavy applications.

I'm not an AI/ML engineer, and I'm not building models. I'm an infrastructure and reliability engineer who finds it fascinating that familiar problems (scheduling, scaling, observability, failure handling) get harder and more interesting when GPUs, large models and unpredictable load enter the picture.

What I'm exploring and learning:

• Reliable AI infrastructure: what "production-ready" means for AI-heavy applications
• Kubernetes for AI workloads: how scheduling and resource management change with specialized hardware
• GPU infrastructure concepts: how GPUs are shared, scheduled, monitored and kept healthy
• AI workload orchestration: running training, batch and inference workloads on shared platforms
• Observability for AI systems: which signals matter beyond CPU and memory (latency, throughput, queueing, saturation)
• Scaling AI applications: capacity planning and autoscaling for expensive, bursty workloads
• Reliability engineering for AI systems: SLOs, failure modes and incident response when the workload is non-traditional
• Infrastructure automation: making AI platforms reproducible rather than hand-built
• AIOps: applying automation and analytics to operations, such as noise reduction, anomaly detection and faster diagnosis
• Cloud-native AI infrastructure: how the cloud-native ecosystem is adapting to AI-heavy demand

The tools will change. The fundamentals (Linux, networking, distributed systems, observability) won't.

• Linux internals: namespaces, cgroups, systemd, networking stack
• Kubernetes internals: scheduler, controllers, networking, storage
• Distributed systems fundamentals: consensus, replication, failure models
• Platform engineering: internal platforms, developer experience, self-service
• GPU infrastructure and AI workload scheduling concepts
• Observability for AI-heavy systems

Roadmap: SRE → Platform Engineering → AI Infrastructure / AI InfraOps

I enjoy talking about Linux, Kubernetes, reliability and distributed systems. I'm also happy to compare notes with anyone exploring AI infrastructure. I write about what I learn on Medium.