Phase 14

Agent Engineering

The core of modern AI engineering. Build agents from first principles.

01The Agent Loop: Observe, Think, ActBuild~60 minutesQuiz 02ReWOO and Plan-and-Execute: Decoupled PlanningBuild~60 minutesQuiz 03Reflexion: Verbal Reinforcement LearningBuild~60 minutesQuiz 04Tree of Thoughts and LATS: Deliberate SearchBuild~75 minutesQuiz 05Self-Refine and CRITIC: Iterative Output ImprovementBuild~60 minutesQuiz 06Tool Use and Function CallingBuild~60 minutesQuiz 07Agent Memory — Virtual Context and Memory PagingBuild~75 minutesQuiz 08Memory Blocks and Sleep-Time ComputeBuild~75 minutesQuiz 09Hybrid Memory: Vector + Graph + KVBuild~75 minutesQuiz 10Skill Libraries and Lifelong Learning (Voyager)Build~75 minutesQuiz 11Planning with HTN and Evolutionary SearchBuild~75 minutesQuiz 12Anthropic's Workflow Patterns: Simple Over ComplexLearn + Build~60 minutesQuiz 13Stateful Graph Orchestration — Durable Execution and CheckpointsLearn + Build~75 minutesQuiz 14The Actor Model for Agents — Async Messages and Typed RuntimesLearn + Build~75 minutesQuiz 15Role-Based Agent Teams — Roles, Tasks, ProcessesLearn + Build~75 minutesQuiz 16OpenAI Agents SDK: Handoffs, Guardrails, TracingLearn + Build~75 minutesQuiz 17The Harness as a Library — Subagents and Session StoreLearn + Build~75 minutesQuiz 18Production Agent Runtimes — Fast Instantiation and Typed WorkflowsLearn~45 minutesQuiz 19Benchmarks: SWE-bench, GAIA, AgentBenchLearn~60 minutesQuiz 20Benchmarks: WebArena and OSWorldLearn~60 minutesQuiz 21Computer Use: Claude, OpenAI CUA, GeminiLearn~60 minutesQuiz 22Voice Agents: Pipecat and LiveKitLearn~60 minutesQuiz 23OpenTelemetry GenAI Semantic ConventionsLearn + Build~60 minutesQuiz 24Agent Observability: Langfuse, Phoenix, OpikLearn~45 minutesQuiz 25Multi-Agent Debate and CollaborationLearn + Build~60 minutesQuiz 26Failure Modes: Why Agents BreakLearn + Build~60 minutesQuiz 27Prompt Injection and the PVE DefenseBuild~75 minutesQuiz 28Orchestration Patterns: Supervisor, Swarm, HierarchicalLearn + Build~60 minutesQuiz 29Production Runtimes: Queue, Event, CronLearn~60 minutesQuiz 30Eval-Driven Agent DevelopmentLearn + Build~60 minutesQuiz 31Agent Workbench Engineering: Why Capable Models Still FailLearn + Build~45 minutesQuiz 32The Minimal Agent WorkbenchBuild~45 minutesQuiz 33Agent Instructions as Executable ConstraintsBuild~50 minutesQuiz 34Repo Memory and Durable StateBuild~60 minutesQuiz 35Initialization Scripts for AgentsBuild~45 minutesQuiz 36Scope Contracts and Task BoundariesBuild~50 minutesQuiz 37Runtime Feedback LoopsBuild~50 minutesQuiz 38Verification GatesBuild~55 minutesQuiz 39Reviewer Agent: Separate Builder from MarkerBuild~55 minutesQuiz 40Multi-Session HandoffBuild~50 minutesQuiz 41The Workbench on a Real RepoBuild~60 minutesQuiz 42Capstone: Ship a Reusable Agent Workbench PackBuild~75 minutesQuiz 43Frame the Task Before the Agent Writes CodeLearn + Build~60 minutesQuiz 44Build an Evidence-Backed Execution PlanLearn + Build~65 minutesQuiz 45Delegate Agent Work with Isolation and Merge ContractsLearn + Build~70 minutesQuiz 46Turn Every Agent Correction into a System ImprovementLearn + Build~65 minutesQuiz 47Define the Outcome Before You Choose the OutputLearn + Build~60 minutesQuiz 48Discover the Workflow People Actually PerformLearn + Build~70 minutesQuiz 49Map Assumptions and Resolve the Riskiest One FirstLearn + Build~65 minutesQuiz 50Choose the Smallest Slice That Can Change the DecisionLearn + Build~65 minutesQuiz 51Write Specifications That Preserve JudgmentLearn + Build~75 minutesQuiz 52Design Success Metrics Before the Result ExistsLearn + Build~70 minutesQuiz 53Choose Prototype, Pilot, or Production DeliberatelyLearn + Build~70 minutesQuiz 54Build a Feedback Ratchet with Ownership and RetirementLearn + Build~75 minutesQuiz

“Stay hungry, stay foolish.”

— Steve Jobs