56

Observability & Tracing with LangSmith and Phoenix

57

Visualizing Token Counts and Latency

58

Automated Evaluation: LLM-as-Judge & Ragas

59

LLM-as-Judge — Part 2: Advanced Evaluation Techniques

60

Evaluating RAG Systems with Ragas

61

Evaluating Agent Trajectories

62

Security & Guardrails: Prompt Injection

63

Output Sanitization

64

LLM Guardrails

65

Safe Sandboxing: Docker & API Key Security

66

Docker Sandboxing — Part 2: API Key Security & Resource Limits

67

API Key Security in Agent Workflows

68

Production Hardening for AI Systems

69

Human-in-the-Loop (HITL) Patterns

70

script_id: day_070_hitl_patterns_part2/multi_stage_approval_pipeline

71

Designing Breakpoints in Agent Systems

72

Injecting Human Feedback into Agent State

73

Capstone — Multi-Agent Content Pipeline with Human Review

Capstone