AI Infra & Serving
PyTorch and vLLM model serving, LiteLLM proxy routing, RunPod & Hyperstack GPU cloud instances, and big-AGI UI integration.
Node Status: CoreOS Online & Rootless
Building Reliable AI & Linux Systems
I'm Juan Karlo "JK" de Guzman—a technical operator focused on production AI infrastructure, rootless Podman containers, financial audit software, and language education.
Platform & Linux
Fedora CoreOS host operations, rootless Podman Quadlets, SELinux policies, Caddy Web Server, and Bunny CDN edge caching.
Systems & Ops
Enterprise financial audit applications, startup launch operations, process flowcharts, SOP creation, and Notion workspaces.
ESL & Leadership
Founder & Instructor at LinguaPath Academy. Coaching candidates for IELTS, TOEFL, LanguageCert ESOL, and Cambridge Business English.
Methodology
Engineering Philosophy
I focus on building systems that hold up under real-world conditions through continuous log inspection, edge-case testing, and systematic root cause isolation.
Infrastructure
Production Server Topology
Select any node below to inspect container mounts, systemd configurations, and active security controls.
Portfolio
Active Projects & Systems
A look at current work across AI infrastructure, financial control applications, platform operations, startup SOPs, and language education.