( https://drive.google.com/drive/folders/1an3DKA3ECzXzAcQ78skbmEM-Oev5Px8m?usp=sharing )

Completed comprehensive AWS training from Internshala, gaining hands-on experience with core AWS services including EC2, S3, Lambda, IAM, AWS databases (RDS), security & management tools, and cost optimization (AWS Billing and Pricing). Developed practical skills in cloud architecture, deployment, and management through a real-world project. Adept at leveraging AWS solutions to optimize infrastructure, enhance scalability, and improve security. Eager to apply cloud computing knowledge to contribute effectively in a dynamic tech environment.

( https://drive.google.com/drive/folders/1LLoBNAnFc9oc-Dn8Qrczs4fvLSbYHE9u?usp=drive_link )

Completed hands-on Cybersecurity Training with a focus on Web Application Vulnerability Assessment using tools like Kali Linux, DVWA, and Burp Suite. Gained practical experience in identifying and exploiting OWASP Top 10 vulnerabilities including SQL Injection, XSS, CSRF, and Command Injection. Performed real-time penetration testing, created Proof-of-Concept (PoC) exploits, and conducted security misconfiguration analysis. Developed strong skills in ethical hacking, threat analysis, and vulnerability reporting.

Breast Cancer Predictor – This project is a small web app that predicts whether a breast cell cluster is benign or malignant using a trained machine learning model. You enter the cytology measurements through sliders, and the app shows the predicted result along with the probability score. It also generates a radar chart so you can compare the mean, standard error, and worst values of the cell features.

Key Features:

• Manually enter cytology measurement values.
• Displays prediction: Benign or Malignant.
• Shows probability for each class.
• Radar chart visualization of important cell features.
• Clean and simple UI (built using Streamlit).

Tech Stack: Python, Streamlit, Plotly, Scikit-Learn, Pickle

Live Link - https://breast-cancer-predictor-by-anuj-arya.streamlit.app/

Virtual Air Mouse – The Virtual AI Mouse is a computer vision-based application that allows users to control their mouse cursor using hand gestures in real-time without touching a physical mouse. Built with Python, OpenCV, MediaPipe, and PyAutoGUI, it uses webcam input to detect hand landmarks and translates specific gestures into mouse actions like movement, clicking, dragging, and scrolling. Key Features:

• 🎯 Cursor Movement: Move the mouse by moving your index finger in front of the webcam.
• 👆 Left Click: Perform a left click using a pinch gesture (thumb + index finger).
• ✌️ Right Click: Trigger a right click using a different gesture (index + middle finger).
• 🤏 Drag and Drop: Pinch and hold to drag files or objects (thumb + index finger).
• 🔄 Scroll Control: Scroll vertically by moving all fingers up/down except thumb.

Tech Stack: Python, OpenCV, MediaPipe, NumPy, PyAutoGUI, AutoPy

DocuMind AI – DocuMind AI is an AI-powered document analysis and interaction tool built with Google Gemini AI and Streamlit. It enables users to upload PDF and TXT documents, generate executive summaries, ask context-grounded questions with precise page citations, extract domain-specific terminology, create visual mind maps, and test their understanding through interactive knowledge challenges. The system combines document parsing, retrieval-augmented generation (RAG), contextual retrieval, and AI-powered analysis to provide reliable, document-grounded responses while maintaining traceability to the original source content. Key features:

• 📄 Document Upload & Parsing: Supports PDF and TXT files with high-fidelity text extraction and page-level chunk tracking.
• 🧠 AI Executive Summaries: Generates concise executive summaries along with important key takeaways from uploaded documents.
• 💬 Grounded Q&A: Ask natural-language questions and receive context-aware answers based on the uploaded document.
• 📑 Precise Page Citations: Provides page references and relevant source snippets to help users verify generated answers.
• 🔍 RAG-Based Retrieval: Retrieves relevant document content before generating answers, improving contextual accuracy and reducing unsupported responses.
• 🗺️ Interactive Mind Maps: Automatically converts document concepts and relationships into visual mind maps using Mermaid diagrams.
• 📚 Glossary & Jargon Extraction: Identifies technical terms, acronyms, and domain-specific terminology and generates their definitions.
• 🎯 Knowledge Challenge: Automatically generates multiple-choice and conceptual questions based on document content and evaluates user responses.
• 🔄 Context-Aware Interaction: Supports document-based conversations and follow-up questions using retrieved contextual information.
• 📤 Multi-Format Export: Export summaries, Q&A history, glossaries, mind maps, and quiz evaluations in Markdown and HTML formats.
• 🌐 Interactive Streamlit Interface: Provides a clean and intuitive interface for uploading documents, analyzing content, asking questions, and exploring generated insights.

Tech Stack: Python, Streamlit, Google Gemini AI, RAG (Retrieval-Augmented Generation), Vector Search, Document Parsing, LangChain

Vulnerability Assessment and Penetration Testing (VAPT) on a Web Application – Successfully completed a Vulnerability Assessment and Penetration Testing (VAPT) project using Kali Linux and DVWA. Explored and exploited various web application vulnerabilities such as SQL Injection, XSS, CSRF, File Inclusion, and Command Injection. Utilized tools like Burp Suite and OWASP ZAP to simulate real-world attack scenarios and documented remediation strategies. Key features:

• Comprehensive Vulnerability Coverage
• Realistic Attack Simulation
• Security Level Adjustment
• Proof of Concept (PoC) Exploits
• Secure Configuration Testing
• Reporting & Documentation

Tech Stack: Kali Linux, DVWA (PHP/MySQL Web App), Burp Suite, PoC scripts