How do you build an agentic AI system that can navigate internal knowledge bases without relying entirely on expensive frontier models? As part of his keynote at Day 1 of Red Hat Summit 2026, CEO Matt Hicks describes how Red Hat engineers moved from basic chatbots to a multi-agent deep research system running directly on Red Hat infrastructure. Watch a detailed breakdown of how Red Hat built an internal agentic system capable of parsing massive bodies of data to return useful insights. Discover the step-by-step process used to replace massive commercial models with open-weight alternatives running on virtual large language model (vLLM) infrastructure. You will learn how small language models (SLMs) can excel at highly specific, isolated tasks within a larger architecture. Resources 🔗 Watch the full Day 1 keynote from Red Hat Summit 2026 → https://www.youtube.com/watch?v=PgMSUGL4N5o 💡 Learn more about enterprise AI use cases and trends → https://www.redhat.com/en/topics/ai 🚀 Read more on our blog → https://www.redhat.com/en/blog/ #RedHat #RHSummit #vLLM #OpenShiftAI #AIAgents #GenerativeAI #OpenWeightModels Keywords Red Hat, Red Hat Summit, AI Agents, Agentic Systems, Open Weight Models, vLLM, Small Language Models, SLM, Chatbots, Hallucination Detection, Multi Agent Systems, Deep Research Agent, Open Source AI, Enterprise AI, OpenShift AI, Matt Hicks

 461      14