>_ ALLISTER ANTOSIKLeveraging automation, machine learning, and high-performance system engineering to build resilient, self-healing platforms.
With extensive experience as a senior developer and DevOps engineer, I bridge the gap between application design and infrastructure reliability.
I design scalable CI/CD pipelines, orchestrate Kubernetes clusters, and implement advanced observability stacks using Prometheus, Grafana, and custom AIOps anomaly detection models.
My approach is rooted in simplicity. I strictly adhere to the KISS (Keep It Simple, Stupid) principle when designing both software and infrastructure architecture.
Complex architectures fail in complex ways. By keeping components decoupled, predictable, and fully automated, we ensure rapid recovery, minimal cognitive load, and maximum system availability.
Recent notes on reliability, tooling, and automation
How I replaced scattered per-client MCP and Skill configuration with MCPM, Skillfish, Mem0, and a symlinked AGENTS.md as a single source of truth across Claude Code and Codex.
Deploy a self-hosted Docker Compose stack that gives Claude Code and Codex a shared OmniRoute model gateway and Mem0 memory over MCP.
Deploy OpenClaw on Unraid with Docker Compose, persistent state, a token-protected Control UI, a CLI companion, and a security-first configuration.
Open to SRE, platform, and reliability consulting, contract work, and interesting collaborations. Drop a line — I read every message.