Agent Substrate: Advancing Kubernetes for Agents at Scale
AI agents spend most of their time idle, waiting on a human prompt, a tool result or a model response. Kubernetes wasnβt built for that. Running one Pod per session holds memory for sessions that arenβt doing anything, and deleting Pods between turns pays a multi-second cold start every time and loses the agentβs state. RL post-training hits the same wall. Agent Substrate, now donated to theβ¦