Agents

How AI agents work — tool use, reasoning, planning, and building them.

• Function Calling & Tool Schemas — Review — First review of function calling concepts — what stuck, what didn’t, and the orchestrator gaps

• Agent Loops & State Management — How agents maintain state across steps, manage growing context, and know when to stop

• ReAct Pattern — Review — Review session on the ReAct pattern — testing recall on the T-A-O loop, failure modes, and what Observation actually means

• Function Calling & Tool Schemas — How LLM agents actually talk to tools — schemas, structured output, and why the orchestrator matters

• ReAct Pattern — How ReAct agents interleave reasoning and acting, failure modes, CoT as computation, and agent pattern tradeoffs