与「RHSummit」相关的搜索结果

At @RedHat, we shifted from proprietary commercial models to small, open-weight models running on vLLM, and it allowed us to successfully scale our internal deep research agent system. Today, open source models handle 85% of our system calls on infrastructure we control! Hear more about our internal journey in my #RHSummit# keynote:

Success in the agentic #AI# era requires a shift from consuming tokens to providing them. @kernelcdub explains why owning your inference infrastructure and routing queries to cost-effective endpoints is the only way to scale without making impossible financial trade-offs. See how to build an agent-ready strategy on Red Hat TV: #RHSummit#