AI Engineering: Agents
Build production-ready AI agents that can autonomously tackle complex workflows, from research to code generation. Learn to architect robust systems that combine LLMs, tools, and memory to ship real-world applications.
Course introduction
Chapter introduction
Deterministic Workflows vs. Agentic Systems
Flexibility, consistency and prompt chaining
Code: transcript analysis
Observability in deterministic chains
Chapter summary
Chapter introduction
Agent autonomy
Components of an Agent
Code: Implementing a Basic Agent Loop in Python
Tracking Agent Components with Weave
Understanding your agent’s traces
OpenAI SDK tracing
Chapter summary
Chapter introduction
Memory types for AI Agents
Online vs offline memory
Code: Implementing Memory for Agents
Observability for Memory Systems in Agents
Observability in Agent Memory Systems
Chapter summary
Chapter introduction
When and Why you should use multiple Agents
Multi-Agent Patterns and Orchestration
Code: Agent Handoffs: Implementing Multi-Agent Systems
OpenAI Agents SDK tracing
Observability in Multi-Agent Systems
Debugging Agent Workflows with Visualization
Chapter summary
Chapter introduction
Why evaluate your agents?
LLM as a judge
Code: simple evals with LLM as a judge
Evaluating Multi-Agent Systems: Dashboards and Metrics
Evaluating Agents with Weave Dashboards
Evaluating Multi-Agent Systems
Comparing Agent Performance
Chapter summary