Joe Fernandes & Brian Stevens, Red Hat | Red Hat Summit 2026
summary
Claims AI becomes a third enterprise pillar and shows concrete plumbing: vLLM GA as the supported inference engine, model-as-a-service to hide infra and meter endpoints, and llm-d distributed inference for scale. Emphasis is on private inference with competitive token economics (move from token consumer toward producer), plus a fledgling agent-ops stack — identity, auth, quota, guardrails, agent tracing and an evaluation hub — all demonstrated in keynote examples and product 3.4.
AI-written summary · about
