Contract Bulk Analyzer

Copyright (c) 2025 Marco De Roni. All rights reserved.
Licensed under the MIT License.

---

Overview

Contract Bulk Analyzer is a Python tool for cross-contract analysis.
Load a folder of PDF and DOCX contracts and get instant aggregate insights: keyword frequency, clause presence rates, metadata extraction, and
group-based comparisons — all exported to Excel and Word automatically.

Built by a Senior Commercial Legal Counsel with EMEA experience in SaaS
enterprise contracting, GDPR compliance, and risk assessment.

Companion project to Contract Scanner
— which generates a detailed per-contract risk report.

---

Features

• 🔍 Keyword frequency — how many contracts contain a given word or phrase
• 📋 Clause presence — % of contracts with/without a specific clause
• 📊 Metadata extraction — parties, governing law, notice period, duration,
  auto-renewal across all contracts in one table
• 🔀 Group comparison — compare clause presence between contract types
  (e.g. vendor vs customer contracts)
• 📊 Excel output — multi-sheet workbook with color-coded results
• 📝 Word output — summary report with tables and divergence highlights
• ⚙️ No coding required — define all queries in plain YAML
• 🔒 PII sanitization — automatically redacts names, dates, emails and other sensitive entities before sending to AI provider, then restores them in the final report
• 🌐 Streamlit web app — browser interface, no command line required
• 📋 Audit log — append-only JSONL log with SHA256 integrity hash
• 🔒 PII redaction report — entities redacted before analysis, restored in report
• 🎯 Confidence score — HIGH/MEDIUM/LOW rating per keyword
• 📄 Clause extraction — extracts exact clause text into dedicated Excel sheet

---

Project Structure

contract-bulk-analyzer/
├── config/
│   ├── queries.example.yaml   # Template — copy and customise
│   └── queries.yaml           # Your queries (excluded from git)
├── contracts/                 # Drop your PDF/DOCX files here (excluded from git)
├── output/                    # Generated Excel + Word reports (excluded from git)
├── analyzer/
│   ├── extractor.py           # PDF and DOCX text extraction
│   ├── keyword_scan.py        # Keyword frequency analysis
│   ├── clause_scan.py         # Clause presence/absence detection
│   ├── metadata_scan.py       # Metadata extraction
│   ├── comparator.py          # Group-based contract comparison
│   └── reporter.py            # Excel and Word report generation
├── main.py                    # Entry point
├── requirements.txt           # Python dependencies
├── LICENSE                    # MIT License
└── README.md

---

Getting Started

1. Clone the repository

git clone https://github.com/marcoderoni/contract-bulk-analyzer.git
cd contract-bulk-analyzer

2. Create and activate virtual environment

python3 -m venv venv
source venv/bin/activate        # Mac/Linux
venv\Scripts\activate           # Windows

3. Install dependencies

pip install -r requirements.txt

4. Set up your queries

cp config/queries.example.yaml config/queries.yaml

Edit config/queries.yaml to define your keywords, clauses, and groups.
The file is excluded from git — your queries stay private.

5. Add contracts

Drop one or more .pdf or .docx files into the contracts/ folder.
The folder is excluded from git — your documents never leave your machine.
You can analyse any number of contracts in a single run.

6. Run

python3 main.py

Results are printed in the terminal with colour coding, and two files
are saved in output/: an Excel workbook and a Word report.

---

How to Customise Queries

All analysis is driven by config/queries.yaml. No code changes needed.

Search for keywords

keywords:
  - "unlimited liability"
  - "auto-renewal"
  - "GDPR"
  - "termination for convenience"

Output: for each keyword, how many contracts contain it and what percentage.

Check clause presence

clauses:
  - name: "Limitation of Liability"
    keywords: ["liability cap", "cumulative liability", "limitation of liability"]
  - name: "Data Protection / GDPR"
    keywords: ["GDPR", "personal data", "data processing"]

Output: per-contract yes/no matrix + % presence across all contracts.

Group contracts for comparison

comparison_groups:
  - name: "Vendor Contracts"
    patterns: ["vendor", "supplier", "MSA"]
  - name: "Customer Contracts"
    patterns: ["customer", "client", "order form"]

Output: clause presence % per group + divergences highlighted
(clauses where groups differ by more than 30%).

Extract metadata

metadata:
  parties_keywords: ["between", "entered into by"]
  governing_law_keywords: ["governed by", "governing law"]
  notice_keywords: ["days' notice", "written notice"]
  duration_keywords: ["term", "duration"]
  renewal_keywords: ["auto-renewal", "automatically renew"]

Output: one row per contract with all extracted fields in Excel.

---

Example Terminal Output

=== Contract Bulk Analyzer ===

📂 Loading contracts...
   ✓ 12 contracts loaded

🔍 Keyword frequency scan...
   unlimited liability          3/12  (25.0%)
   auto-renewal                 9/12  (75.0%)
   GDPR                        11/12  (91.7%)

📋 Clause presence scan...
   Limitation of Liability     100.0%
   Data Protection / GDPR       91.7%  ← missing in: contract_03.pdf
   Force Majeure                41.7%

📊 Metadata extraction...
   ✓ Metadata extracted from 12 contracts

🔀 Group comparison...
   ⚠️  2 significant divergences found:
   Force Majeure: Customer Contracts 80% vs Vendor Contracts 20%

📝 Generating reports...
✅ Done!
   📊 Excel: output/bulk_analysis_20250323_1700.xlsx
   📝 Word:  output/bulk_report_20250323_1700.docx

---

Privacy & Confidentiality

• contracts/ is excluded from git — your documents never leave your machine
• config/queries.yaml is excluded from git — your queries stay private
• All processing is local — no data is sent to external APIs
• PII redaction via Microsoft Presidio — 900+ entities automatically anonymised per batch

---

Requirements

• Python 3.9+
• pdfplumber
• python-docx
• pyyaml
• openpyxl
• colorama
• tqdm

---

Related Project

Contract Scanner
Single-contract review tool with R/Y/G risk scoring and detailed Word report.

---

Author

Marco De Roni
Senior Commercial Legal Counsel | EMEA
LinkedIn · GitHub

---

Demo

---

License

MIT License — see LICENSE for details.