I build cloud platforms and the observability around them: OpenTelemetry tracing for LLM and agent workloads, zero-trust AWS blueprints, and Python services for an OpenStack platform of 50K+ instances.
How I orchestrate autonomous subagents in the terminal (Hermes, SQLite kanban, LLM fallback)
How I retired R$200/month in cloud LLMs with a local hybrid pipeline





