# sandesh-hase/langgraph-tutorial architecture

> Architecture diagram of the GitHub repository sandesh-hase/langgraph-tutorial, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-07). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/sandesh-hase/langgraph-tutorial
- Repository: https://github.com/sandesh-hase/langgraph-tutorial

## Overview

This repository is a LangGraph tutorial series, with a Streamlit warranty-claim fraud detector as its main end-to-end application and a separate Studio shopping-agent example. The principal workflow is CSV upload, policy review, fraud scoring, evidence collection, adjudication, and result inspection/export. The notebooks cover graph fundamentals and interaction patterns; the README also describes visualization, execution tracing, and LangSmith-based observability. The graph emphasizes relationships visible in the README or sampled source; unsampled notebook internals and Studio hosting behavior are not inferred.

Main entry points: [`PROJECT-warranty-fraud-detector/main.py`](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py) (Claim Processor), [`PROJECT-warranty-fraud-detector/app.py`](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/app.py) (Claims UI), [`05-studio/agent.py`](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py) (Shopping Agent), [`PROJECT-warranty-fraud-detector/data/AutoDrive_Warranty_Policy_2025.pdf`](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/data/AutoDrive_Warranty_Policy_2025.pdf) (Warranty Policy)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_tutorials["LangGraph Tutorials"]
  node_fundamentals["Graph Fundamentals"]
  node_conditional["Conditional Graphs"]
  node_memory["Memory Management"]
  node_streaming["Streaming &amp; Interrupts"]
  node_human_loop["Human Feedback"]
  node_state_edit["State Editing"]
end

subgraph group_warranty["Warranty Detector"]
  node_streamlit["Claims UI<br/>[app.py]"]
  node_processor["Claim Processor<br/>[main.py]"]
  node_policy["Policy Review<br/>[main.py]"]
  node_fraud["Fraud Scoring<br/>[main.py]"]
  node_evidence["Evidence Collection<br/>[main.py]"]
  node_adjudication["Claim Decision<br/>[main.py]"]
  node_policy_pdf["Warranty Policy"]
end

subgraph group_shopping["Shopping Agent"]
  node_shopping_ui["Shopping Agent<br/>[agent.py]"]
  node_shopping_model["Tool-Bound Model<br/>[agent.py]"]
  node_shopping_tools["Shopping Tools<br/>[agent.py]"]
end

subgraph group_observability["Tracing &amp; Inspection"]
  node_traces["Execution Traces<br/>[main.py]"]
  node_langsmith["LangSmith"]
end

node_claimant(("Claims User"))
node_azure["Azure OpenAI"]

node_claimant -->|"uploads claims"| node_streamlit
node_streamlit -->|"processes claims"| node_processor
node_processor -->|"runs review"| node_policy
node_processor -->|"runs scoring"| node_fraud
node_processor -->|"collects evidence"| node_evidence
node_processor -->|"requests decision"| node_adjudication
node_policy -->|"loads policy"| node_policy_pdf
node_policy -->|"invokes model"| node_azure
node_fraud -->|"invokes model"| node_azure
node_evidence -->|"invokes model"| node_azure
node_adjudication -->|"invokes model"| node_azure
node_processor -->|"records agent steps"| node_traces
node_streamlit -->|"shows and exports"| node_claimant
node_shopping_ui -->|"invokes model"| node_shopping_model
node_shopping_model -->|"dispatches tool calls"| node_shopping_tools
node_shopping_tools -->|"returns tool results"| node_shopping_ui
node_traces -.->|"traces runs"| node_langsmith

click node_streamlit "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/app.py"
click node_processor "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py"
click node_policy "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py"
click node_fraud "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py"
click node_evidence "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py"
click node_adjudication "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py"
click node_policy_pdf "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/data/AutoDrive_Warranty_Policy_2025.pdf"
click node_fundamentals "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/01-LangGraph_Basics.ipynb"
click node_conditional "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/02-Conditional-Graph.ipynb"
click node_memory "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/03-LangGraph-Memory-Management-travel_agent_explanation.ipynb"
click node_streaming "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/04-streaming-interruption.ipynb"
click node_human_loop "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-Human-in-the-loop.ipynb"
click node_state_edit "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/06-edit-state-human-feedback.ipynb"
click node_shopping_ui "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py"
click node_shopping_model "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py"
click node_shopping_tools "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py"
click node_traces "https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_fundamentals,node_conditional,node_memory,node_streaming,node_human_loop,node_state_edit,node_claimant toneBlue
class node_streamlit,node_processor,node_policy,node_fraud,node_evidence,node_adjudication,node_policy_pdf toneAmber
class node_shopping_ui,node_shopping_model,node_shopping_tools toneMint
class node_traces,node_langsmith toneRose
class node_azure toneIndigo
```

## Components

### LangGraph Tutorials

- [Graph Fundamentals](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/01-LangGraph_Basics.ipynb): `01-LangGraph_Basics.ipynb`
- [Conditional Graphs](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/02-Conditional-Graph.ipynb): `02-Conditional-Graph.ipynb`
- [Memory Management](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/03-LangGraph-Memory-Management-travel_agent_explanation.ipynb): `03-LangGraph-Memory-Management-travel_agent_explanation.ipynb`
- [Streaming & Interrupts](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/04-streaming-interruption.ipynb): `04-streaming-interruption.ipynb`
- [Human Feedback](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-Human-in-the-loop.ipynb): `05-Human-in-the-loop.ipynb`
- [State Editing](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/06-edit-state-human-feedback.ipynb): `06-edit-state-human-feedback.ipynb`

### Warranty Detector

- [Claims UI](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/app.py): `PROJECT-warranty-fraud-detector/app.py`
- [Claim Processor](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py): `PROJECT-warranty-fraud-detector/main.py`
- [Policy Review](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py): `PROJECT-warranty-fraud-detector/main.py`
- [Fraud Scoring](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py): `PROJECT-warranty-fraud-detector/main.py`
- [Evidence Collection](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py): `PROJECT-warranty-fraud-detector/main.py`
- [Claim Decision](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py): `PROJECT-warranty-fraud-detector/main.py`
- [Warranty Policy](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/data/AutoDrive_Warranty_Policy_2025.pdf): `PROJECT-warranty-fraud-detector/data/AutoDrive_Warranty_Policy_2025.pdf`

### Shopping Agent

- [Shopping Agent](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py): `05-studio/agent.py`
- [Tool-Bound Model](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py): `05-studio/agent.py`
- [Shopping Tools](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/05-studio/agent.py): `05-studio/agent.py`

### Tracing & Inspection

- [Execution Traces](https://github.com/sandesh-hase/langgraph-tutorial/blob/main/PROJECT-warranty-fraud-detector/main.py): `PROJECT-warranty-fraud-detector/main.py`
- **LangSmith**

### Other components

- **Claims User**
- **Azure OpenAI**

## Connections

- Claims User → Claims UI: uploads claims (evidence: `PROJECT-warranty-fraud-detector/app.py`)
- Claims UI → Claim Processor: processes claims (evidence: `PROJECT-warranty-fraud-detector/app.py`)
- Claim Processor → Policy Review: runs review (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Claim Processor → Fraud Scoring: runs scoring (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Claim Processor → Evidence Collection: collects evidence (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Claim Processor → Claim Decision: requests decision (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Policy Review → Warranty Policy: loads policy (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Policy Review → Azure OpenAI: invokes model (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Fraud Scoring → Azure OpenAI: invokes model (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Evidence Collection → Azure OpenAI: invokes model (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Claim Decision → Azure OpenAI: invokes model (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Claim Processor → Execution Traces: records agent steps (evidence: `PROJECT-warranty-fraud-detector/main.py`)
- Claims UI → Claims User: shows and exports (evidence: `PROJECT-warranty-fraud-detector/app.py`)
- Shopping Agent → Tool-Bound Model: invokes model (evidence: `05-studio/agent.py`)
- Tool-Bound Model → Shopping Tools: dispatches tool calls (evidence: `05-studio/agent.py`)
- Shopping Tools → Shopping Agent: returns tool results (evidence: `05-studio/agent.py`)
- Execution Traces → LangSmith: traces runs (evidence: `README.md`)
