Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn to build and deploy intelligent AI agent systems that go beyond simple chatbots and prompt chaining in this 16-minute conference talk by Microsoft Principal Applied Scientist Michael Albada. Discover the core design principles behind autonomous agentic systems, starting with a clear definition of what constitutes an AI agent versus traditional language model applications. Explore three fundamental components: tool use and function calling capabilities, orchestration patterns including chains, trees, and agentic workflows, and multi-agent system architectures. Master the architectural trade-offs involved in coordinating multiple agents to handle complex workflows with minimal human oversight. Avoid common development pitfalls including insufficient evaluation practices, lack of observability, tool integration issues, and unnecessary complexity. Examine specific evaluation tools and methodologies for assessing agent performance, while understanding critical security and safety considerations for production deployments. Gain actionable insights from real-world implementations and latest research to design tailored solutions that enhance efficiency and innovation, whether building single-agent systems or scaling up to complex multi-agent architectures.
