01 / Platform engineering for AI systems
Reliable
systems for
unreliable
frontiers.
Hans Lao builds the secure, observable foundations that make AI useful in production: infrastructure-as-code, Kubernetes, model serving, and agent governance.
Active systems / topologyPlatform
engineeringIaC
TerraformAI agents
MCPObservability
SRESecurity
IAM
engineeringIaC
TerraformAI agents
MCPObservability
SRESecurity
IAM
Uptime99%infrastructure uptime maintained
Platform53services in a 100% IaC homelab
Research12axes in the Rubicon benchmark
Scale4K+enterprise endpoints operated
02 / Selected systems
Evidence, not
technology lists.
01100% IaC Homelab PlatformA private cloud built entirely as code — Proxmox, Terraform, Traefik, and 50+ services with secrets management and observability.↗02notification-state PlatformA production incident-notification system: FastAPI + SQLite dashboard, n8n routing, ntfy push, six systemd-timer producers, and encrypted backups.↗03Hermes Agent + TaskNotes SkillA persistent AI agent (DeepSeek backend, Telegram interface, memory) extended with a custom skill that manages tasks across Obsidian vaults.↗
03 / Active research
Can an AI model be trusted with production infrastructure?
At Segen, I work on reliable AI platforms. Independently, I am developing Rubicon: a safety-first, 12-axis benchmark for judging AI infrastructure work.
Read the research note ↗JUDGEMENT
OVER
COMPLIANCE.
OVER
COMPLIANCE.