🔒 SecureLens AI

AI-Powered Cybersecurity & Compliance Document Analyzer

SecureLens AI is an intelligent cybersecurity assistant that automatically analyzes uploaded documents for sensitive information, evaluates compliance risks, generates AI-powered security reports, performs document masking, supports OCR for image-based documents, and enables natural language interaction through Retrieval-Augmented Generation (RAG).

The system combines traditional pattern matching, OCR, vector search, and Generative AI to help organizations identify confidential information and improve document security before sharing or storing files.

🚀 Live Demo

Public Deployment:
👉 https://securelens-ai-production.up.railway.app

🎯 Project Objective

Organizations frequently share resumes, contracts, reports, invoices, identity documents, and confidential files without realizing that they may contain personally identifiable information (PII) or sensitive credentials.

SecureLens AI aims to:

Detect confidential and sensitive information
Assess cybersecurity and compliance risks
Generate AI-powered compliance reports
Protect sensitive information through masking
Allow users to ask questions about uploaded documents using AI
Support both text-based and scanned documents

✨ Features

📄 Multi-format Document Support

PDF Documents
TXT Files
CSV Files
Image Files (PNG, JPG, JPEG) using OCR

🔍 Sensitive Data Detection

Automatically detects:

Email Addresses
Phone Numbers
PAN Numbers
Aadhaar Numbers
Credit Card Numbers
Employee IDs
Bank Account Numbers
IFSC Codes
API Keys
GitHub Tokens
JWT Tokens
Passwords
Confidential Keywords

🤖 AI Compliance Report

Google Gemini generates:

Compliance observations
Security risks
Recommended actions
Executive summary

📊 Risk Classification

Documents are automatically classified into:

Low Risk
Medium Risk
High Risk

based on detected sensitive information.

💬 AI Compliance Chatbot (RAG)

Users can ask questions like:

Summarize this document
How many emails are present?
List all sensitive information
What security risks exist?
What should I improve?

The chatbot combines:

Retrieval-Augmented Generation (RAG)
FAISS Vector Database
Sentence Transformers
Google Gemini

🔐 Data Masking

Automatically masks sensitive information such as:

Email:
john****@gmail.com

Phone:
98******21

PAN:
ABCDE****F

🖼 OCR Support

Image-based documents are processed using EasyOCR.

Example:

Aadhaar Card
Identity Cards
Scanned Documents

💼 Use Cases

Resume Security Analysis
Employee Document Screening
Compliance Audits
Secure Document Sharing
HR Verification
Identity Document Analysis
Cybersecurity Assessments
Data Privacy Compliance

✅ Advantages

Detects sensitive information automatically
Reduces accidental data exposure
AI-generated cybersecurity insights
Supports scanned image documents
Interactive AI assistant
Easy-to-use web interface
Fast semantic document search
Privacy-aware document masking

🏗 System Architecture

                User
                  │
                  ▼
        Upload Document
                  │
                  ▼
        Document Parser
       (PDF / TXT / CSV / OCR)
                  │
                  ▼
     Sensitive Data Detection
      (Regex + Keywords)
                  │
                  ▼
      Risk Classification
                  │
                  ▼
       AI Compliance Report
        (Google Gemini)
                  │
                  ▼
      FAISS Vector Database
                  │
                  ▼
    RAG AI Compliance Chatbot
                  │
                  ▼
        Data Masking Module
                  │
                  ▼
          User Dashboard

🧠 AI / ML Approach Used

SecureLens AI combines multiple AI techniques to provide intelligent document analysis.

1. OCR

EasyOCR extracts text from image-based documents.

2. Sensitive Data Detection

Regex-based detection identifies structured information including:

Emails
Phone Numbers
PAN
Aadhaar
Credit Cards
API Keys
Passwords

Keyword-based detection identifies confidential terminology.

3. Risk Assessment

Each detected entity is assigned a weighted score.

Overall document risk is classified into:

Low
Medium
High

4. Retrieval-Augmented Generation (RAG)

The uploaded document is:

Split into chunks
Embedded using Sentence Transformers
Stored in a FAISS vector database

When the user asks a question:

Relevant document chunks are retrieved
Sent to Google Gemini
Gemini generates context-aware responses

5. AI Compliance Report

Google Gemini analyzes:

Compliance observations
Security risks
Recommendations
Executive summary

💻 Tech Stack

Frontend

HTML5
CSS3
JavaScript
Tailwind CSS

Backend

Flask
Python

AI / ML

Google Gemini API
Sentence Transformers
FAISS
EasyOCR

Libraries

pdfplumber
pandas
NumPy
python-dotenv

Deployment

Railway

📸 Output Screens

Include screenshots of:

Home Page

screenshots/home.png

Document Upload

screenshots/upload.png

Sensitive Information Detection

screenshots/detection.png

AI Compliance Report

screenshots/report.png

AI Chatbot

screenshots/chatbot.png

Data Masking

screenshots/masking.png

⚠ Challenges Faced

During development, several technical challenges were encountered:

Extracting text from scanned image documents.
Improving sensitive data detection accuracy while minimizing false positives.
Implementing Retrieval-Augmented Generation (RAG) for document-specific question answering.
Managing large AI dependencies such as EasyOCR and Sentence Transformers.
Optimizing deployment due to memory constraints on cloud hosting platforms.
Designing a responsive and intuitive cybersecurity dashboard.
Ensuring AI responses remain context-aware and document-specific.

🚀 Future Improvements

Multi-document support
PDF download of masked documents
Automatic document redaction
Compliance support for GDPR, HIPAA, ISO 27001, PCI-DSS
User authentication
Document history dashboard
Batch document analysis
Cloud storage integration
Role-based access control
Real-time compliance monitoring
AI-powered policy recommendation engine

⚙ Installation & Local Setup

1. Clone the Repository

git clone https://github.com/ManvithaPola/SecureLens-AI.git

cd SecureLens-AI

2. Create Virtual Environment

Windows

python -m venv venv

Activate

venv\Scripts\activate

3. Install Dependencies

pip install -r requirements.txt

4. Create Gemini API Key

Visit:

https://aistudio.google.com/app/apikey

Create a free Gemini API Key.

5. Create a `.env` File

Inside the project root create:

.env

Add:

GEMINI_API_KEY=YOUR_GEMINI_API_KEY

6. Run the Application

python app.py

The application will start at:

http://127.0.0.1:5000

📁 Project Structure

SecureLens-AI/

│── app.py
│── requirements.txt
│── README.md
│── .env
│
├── modules/
│      parser.py
│      detector.py
│      classifier.py
│      chat.py
│      rag.py
│      summarizer.py
│      masker.py
│      keyword_detector.py
│      ocr.py
│
├── templates/
│
├── uploads/
│
└── screenshots/

👩‍💻 Author

Manvitha Pola

Artificial Intelligence & Machine Learning Engineer

📜 License

This project is intended for educational and research purposes.

Name		Name	Last commit message	Last commit date
modules		modules
screenshots		screenshots
templates		templates
uploads		uploads
.gitignore		.gitignore
README.md		README.md
app.py		app.py
architecture.png		architecture.png
requirements.txt		requirements.txt
test_rag.py		test_rag.py

Latest commit

History

Folders and files

Repository files navigation