Completed
Why LangGraph: State Machines for Branching, Retries, and Traceability
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Orchestrating Agentic State Machines with LangGraph
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro: Orchestrating Agent State Machines for Clinical Documentation
- 2 What “Agentic” Means Engineering View: Nodes, State, and Routing
- 3 The Clinical Scribe Use Case + 4-Agent Pipeline Overview SOAP
- 4 Why This Matters: Documentation Burden, Burnout, and System Impact
- 5 Why Not One Big Prompt: Accuracy, Auditability, and Debuggable Workflows
- 6 Why Clinical Notes Are Hard: Messy Dialog, Negation, Time, Uncertainty
- 7 What We’re Building: Inputs/Outputs, Pipeline, and Two Key Design Choices
- 8 Why LangGraph: State Machines for Branching, Retries, and Traceability
- 9 LangGraph Vocabulary Crash Course: State, Nodes, Edges, Routers, Cycles
- 10 The Shared MedicalScribeState: Audit Logs, Errors, and Testability
- 11 Schemas & Structured Extraction: Pydantic Models as an API Contract
- 12 Node Implementation Template + Transcription Agent Responsibilities
- 13 Extraction Agent Deep Dive: Validation Retries and Negation Handling
- 14 Negation, temporal qualifiers & deterministic normalization in extraction
- 15 From structured encounter to SOAP note: summarization constraints & editable sections
- 16 Reducing hallucinations: structured inputs, no-new-facts prompts & low temperature
- 17 Consistency agent: catching omissions, contradictions & invented facts
- 18 Consistency checks taxonomy: hard rules vs semantic judging vs safety scans
- 19 Graph orchestration with LangGraph: explicit routing, safe failures & debugging
- 20 End-to-end demo walkthrough: transcription → extraction → SOAP → consistency
- 21 Retrieval grounding: mapping diagnoses to ICD-10/SNOMED as a pipeline node
- 22 Measuring quality per agent: datasets, metrics & iteration strategy
- 23 Production readiness: PHI, redaction, prompt injection, monitoring & human review
- 24 Setup, key takeaways & next steps feedback loop, better grounding
