AI Logistics Report Documentation & Validation

An autonomous n8n workflow built to automate the manual validation of daily shipment records and instantly generate dynamic shipping manifest PDFs to prevent dispatch delays.

🎯 Client Goal

To eliminate manual data entry for a fast-growing logistics company, ensuring zero dispatch errors while processing high volumes of unstructured shipment data.

🛠️ Tech Stack

• Workflow Engine: n8n
• AI/LLM: OpenAI API (Agentic Validation & Structured Output Parsing)
• Document Generation: PDFMonkey API
• Database & Storage: Google Sheets, Google Drive
• Communication: Gmail API

graph LR
    %% Trigger and Loop
    Trigger([Google Sheets Trigger<br/>⚡ rowAdded]) --> Loop(Loop Over Items)

    %% AI Agent Node & Its Dependencies
    Model((OpenAI Chat Model)) -.->|Model| Agent[Validate Shipment details]
    Context[(Carrier Details<br/>Read Sheet)] -.->|Data| Agent
    Parser{{Structured Output Parser}} -.->|Parser| Agent

    Loop -->|loop| Agent
    Agent --> Cond{If Validate}

    %% True Branch (Validation Successful)
    Cond -->|true| SetData[Set Shipment Data<br/>Manual]
    SetData --> PDFGen[PDF Generate<br/>PDFMonkey]
    PDFGen --> GetPDF[Get PDF]
    GetPDF --> EmailSender[Send Shipment PDF<br/>To Sender]
    EmailSender --> EmailReceiver[Send Shipment PDF<br/>To Receiver]
    EmailReceiver --> Download[Download File]
    Download --> UploadDrive[Upload file<br/>Google Drive]
    UploadDrive --> UpdateSheet[Append or update row<br/>in sheet]
    
    %% Loop back to next item
    UpdateSheet -->|Next Item| Loop

    %% False Branch (Validation Failed / Mismatch)
    Cond -->|false| MismatchEmail[Send Mismatch Items<br/>Details via Gmail]
    MismatchEmail --> UpdateState[Update State<br/>Append/Update Sheet]
    
    %% Loop back to next item from false branch
    UpdateState -->|Next Item| Loop

    %% End of Loop
    Loop -.->|done| End([End Workflow])

    %% Styling to match node types (Optional but looks good)
    classDef trigger fill:#0b8a5b,stroke:#065f3d,stroke-width:2px,color:#fff;
    classDef ai fill:#7c3aed,stroke:#5b21b6,stroke-width:2px,color:#fff;
    classDef logic fill:#0ea5e9,stroke:#0369a1,stroke-width:2px,color:#fff;
    classDef external fill:#f43f5e,stroke:#be123c,stroke-width:2px,color:#fff;
    classDef sheets fill:#10b981,stroke:#047857,stroke-width:2px,color:#fff;

    class Trigger,UpdateSheet,UpdateState,Context sheets;
    class Agent,Model,Parser ai;
    class Loop,Cond,SetData logic;
    class PDFGen,GetPDF,EmailSender,EmailReceiver,MismatchEmail,Download,UploadDrive external;

Loading

🧠 How It Works

1. Trigger: Workflow initiates upon a new entry in Google Sheets.
2. AI Validation: An OpenAI agent cross-references unstructured raw shipment data with carrier databases.
3. Conditional Routing:
   • Match (True): Generates a dynamic PDF via PDFMonkey, uploads to Drive, and emails the sender/receiver.
   • Mismatch (False): Instantly sends an alert containing the mismatched data for manual review.

📈 Impact & Results

• Time Saved: 40+ hours per week for the operations team.
• Accuracy: Reduced dispatch errors to zero.
• Scalability: Allowed the client to scale shipment volume 10x seamlessly without adding operational headcount.

---

Developed by Uba Chan