Sessions
.NET Agents & Hybrid AI - Bridging Local and Cloud Models with Agent Framework
Cloud LLMs are extremely powerful but pose data privacy risks and cause high operational costs. Local (and on-device) SLMs keep your sensitive data secure and run “for free”, but lack deep reasoning skills. Choosing just one means compromising. This session explores the blueprints for collaborative systems - we will look at the latest research patterns related to local-remote model orchestration and then leverage Agent Framework in .NET to create intelligent systems that dynamically route and cascade tasks between local models and deep-reasoning cloud LLMs.
