Too Much Lock-in for Too Little Gain - Agent Frameworks Are a Dead-End

Learn how to build agentic AI systems without getting locked into specific frameworks in this conference talk by Valliappa Lakshmanan from MLOps.community. Discover why agent frameworks may be creating unnecessary constraints and explore an alternative architecture using simple, composable GenAI patterns combined with off-the-shelf monitoring and logging tools. Understand how to accelerate development of production-quality agentic systems while maintaining flexibility across different LLMs, cloud platforms, and frameworks. Explore practical design patterns from Lakshmanan's GenAI design patterns book that enable you to build scalable AI agents without vendor lock-in. Gain insights from an industry expert who previously served as Director for Data Analytics and AI Solutions at Google Cloud, co-founded Google's Advanced Solutions Lab, and currently works as an operating executive helping portfolio companies implement data-driven AI innovation strategies.