🤖 AutoOps AI — Document to Dashboard Automation System

A Python-powered AI automation engine that transforms business documents (invoices, contracts, bank statements, reports) into structured data, insights, and interactive dashboards — automatically.

---

✨ Features

#	Feature	Description
1	Document Upload	Upload PDF invoices, contracts, bank statements, reports
2	Automated Extraction	Extract structured data using pdfplumber + regex patterns
3	LLM Summarization	OpenAI-powered executive summaries, risk flags, action items
4	Structured JSON Output	Clean, schema-consistent JSON for every document
5	Interactive Dashboard	Streamlit dashboard with charts, risk alerts, spending trends
6	Automation Layer	Auto-save to CSV/SQLite, PDF report generation, email notifications
7	Zapier Integration	Webhook API to connect with 6000+ apps (Slack, Sheets, QuickBooks)

---

📁 Project Structure

Automation_project/
├── config.py               # Central configuration
├── extractor.py             # PDF text & field extraction
├── summarizer.py            # OpenAI LLM summarization
├── pipeline.py              # End-to-end processing pipeline
├── database.py              # SQLite storage layer
├── automation.py            # CSV export, PDF reports, email
├── dashboard.py             # Streamlit interactive dashboard
├── main.py                  # CLI entry point
├── generate_samples.py      # Sample PDF generator for testing
├── requirements.txt         # Python dependencies
├── .env.example             # Environment variables template
├── .gitignore
├── samples/                 # Generated sample PDFs
├── uploads/                 # Uploaded documents
└── outputs/
    ├── json/                # Structured JSON outputs
    ├── csv/                 # CSV exports
    └── reports/             # Generated PDF reports

---

🚀 Quick Start

1. Install Dependencies

cd Automation_project
pip install -r requirements.txt

2. Configure Environment

# Copy the template
cp .env.example .env

# Edit .env and add your OpenAI API key
# OPENAI_API_KEY=sk-your-key-here

Note: The system works without an API key — it will skip LLM analysis and only use regex extraction.

3. Generate Sample PDFs

python generate_samples.py

This creates 4 sample PDFs in samples/:

• sample_invoice.pdf
• sample_contract.pdf
• sample_bank_statement.pdf
• sample_quarterly_report.pdf

4. Launch the Dashboard

streamlit run dashboard.py

Open http://localhost:8501 in your browser.

---

💻 CLI Usage

# Process a single PDF
python main.py process samples/sample_invoice.pdf

# Process without LLM (regex-only)
python main.py process samples/sample_invoice.pdf --no-llm

# Batch process all PDFs in a folder
python main.py batch samples/

# Export all records to CSV
python main.py export-csv

# Generate summary PDF report
python main.py export-pdf

# View dashboard stats in terminal
python main.py stats

# Launch Streamlit dashboard
python main.py dashboard

---

🔄 Processing Pipeline

 PDF Upload
     │
     ▼
 ┌─────────────────┐
 │  PDF Extraction  │  pdfplumber → raw text + tables
 │  + Regex Fields  │  regex patterns → invoice #, date, amount, vendor...
 └────────┬────────┘
          │
          ▼
 ┌─────────────────┐
 │  Document Type   │  Heuristic keyword scoring
 │  Detection       │  invoice / contract / bank_statement / report / email
 └────────┬────────┘
          │
          ▼
 ┌─────────────────┐
 │  LLM Analysis   │  OpenAI API (gpt-4o-mini)
 │  (if enabled)    │  → Summary, risk flags, action items, structured JSON
 └────────┬────────┘
          │
          ▼
 ┌─────────────────┐
 │  Merge & Store   │  Regex + LLM fields → canonical output
 │                  │  → SQLite DB + JSON file
 └────────┬────────┘
          │
          ▼
 ┌─────────────────┐
 │  Automation      │  → CSV append
 │  Layer           │  → PDF report generation
 │                  │  → Email notification (optional)
 └────────┬────────┘
          │
          ▼
 ┌─────────────────┐
 │  Dashboard       │  Charts, risk alerts, spending trends
 │  (Streamlit)     │  Search, export, download reports
 └─────────────────┘

---

📊 Dashboard Tabs

📤 Upload & Process

• Drag & drop PDFs
• Toggle AI analysis on/off
• View extraction results, risk badges, summaries
• Download individual PDF reports

📊 Analytics Dashboard

• Total documents & total amount metrics
• Document type distribution (pie chart)
• Risk level distribution (bar chart)
• Monthly spending trend (line chart)
• Risk alerts & pending payments tables

📋 Documents

• Search across all processed documents
• Sortable data table
• Detail view with full JSON & action items

📥 Export & Reports

• One-click CSV export
• Summary PDF report generation
• Download previous exports

---

📄 Sample JSON Output

{
  "file": "sample_invoice.pdf",
  "document_type": "invoice",
  "invoice_number": "INV-2026-0042",
  "vendor": "Acme Solutions Ltd.",
  "date": "February 20, 2026",
  "due_date": "March 22, 2026",
  "total_amount": 11714.75,
  "currency": "USD",
  "tax_amount": "$917.75",
  "risk_flag": "low",
  "risk_reason": "Late payment fee clause present",
  "summary": "Invoice from Acme Solutions for cloud hosting, API development, and support services totaling $11,714.75 due March 22, 2026.",
  "action_items": [
    "Schedule payment before March 22 to avoid late fees",
    "Verify API integration hours against project tracker",
    "Confirm SSL certificate count matches deployed domains"
  ],
  "line_items": [
    {
      "description": "Cloud Hosting Services",
      "quantity": "1",
      "unit_price": "$2,500.00",
      "amount": "$2,500.00"
    },
    {
      "description": "API Integration Development",
      "quantity": "40 hrs",
      "unit_price": "$150.00",
      "amount": "$6,000.00"
    }
  ]
}

---

🔗 Zapier Integration

Connect AutoOps AI with 6000+ apps to automate your entire workflow.

Quick Setup

1. Start the webhook server:

   python webhook_api.py

2. Expose your local server for testing:

   # Install ngrok from https://ngrok.com/download
   ngrok http 8000

3. Configure your Zapier Zap:

   • Trigger: Gmail, Dropbox, Slack, etc.
   • Action: Webhooks by Zapier → POST
   • URL: https://your-ngrok-url.ngrok.io/process
   • Headers: X-API-Key: your-secret-api-key-here

Example Workflows

📧 Gmail → AutoOps → Slack
Automatically process invoice attachments and notify your team in Slack.

☁️ Dropbox → AutoOps → Google Sheets
Watch a Dropbox folder and add processed data to a tracking spreadsheet.

💰 Email → AutoOps → QuickBooks
Extract invoice data and create entries in your accounting software.

🚨 High-Risk Alert
Send SMS and create Trello cards when contracts have risk flags.

📖 Read the full Zapier Integration Guide →

---

⚙️ Configuration

Edit config.py or .env to customize:

Setting	Default	Description
OPENAI_MODEL	gpt-4o-mini	OpenAI model to use
TEMPERATURE	0.2	LLM creativity (lower = more deterministic)
MAX_TOKENS	2048	Max response tokens
SMTP_HOST	smtp.gmail.com	Email server for notifications

---

🛠 Tech Stack

• Python 3.11+
• pdfplumber — PDF text & table extraction
• OpenAI API — LLM summarization & structuring
• Streamlit — Interactive web dashboard
• Plotly — Charts & visualizations
• SQLite — Lightweight database
• FPDF2 — PDF report generation
• pandas — Data manipulation

---

📬 Email Notifications (Optional)

To enable email alerts after processing:

1. Set SMTP credentials in .env
2. Enable "Auto-email notification" in the dashboard sidebar
3. For Gmail, use an App Password

---

📜 License

MIT License — free for personal and commercial use.