Gabriel Cosi

Software Engineer

Systems Operational

Running LLMs at home

Building an on-demand inference box.

LLM Usage

Tokens (7d)63,378 tokens

Top Model (7d)deepseek-v4-flash

Cluster Metrics

CPU Load32.1%

Memory42.1%

Active Pods197

Uptime28d

SLA (7d)99.97%

Global P99 (7d)89 ms