Projects

Cyber Threat Intelligence Observatory

• Deployed OpenCTI on the cloud using Docker and OpenStack, enabling Cyber Threat Intelligence for Digital Public Infrastructure(DPI).
• Integrated RSS feeds from key sources to enrich threat intelligence with real-time insights on DPI-related cyber risks.
• Transitioning analysis to Microsoft Sentinel for advanced security monitoring and threat detection.

Privacy-Enhancing Technologies in Zero Trust Network

• Focused on biometric security on Android devices using Trusted Execution Environment (TEE).
• Contributed to this project by enhancing biometric security on Android devices through the integration of TEE. This project focused on leveraging TEE to secure biometric data, ensuring that sensitive information like fingerprints and facial recognition data are stored and processed in a highly secure, isolated environment.
• Developed a Flask server to emulate server-side code, enabling efficient testing and deployment. Additionally, developed an Android app with device attestation performed to check for tampering and verify the authenticity of the app.

Cloud & Virtualization Security Hardening

• Evaluated and hardened a prototype containerized cloud environment, reducing attack surfaces and implementing robust security policies to protect against vulnerabilities.
• Refined Dockerfiles and generated hardened images to enhance the security posture of the containerized application by configuring AppArmor and SecComp profiles.
• Determined and implemented runtime commands for secure container deployment, ensuring proper functionality and adherence to security best practices.
• Conducted security audits using Docker Bench Security and Grype to ensure compliance.

B.Tech Projects

Chatbot Implementation using Deep Natural Language Processing

📅 April 2019
Implemented a chatbot using a Seq2Seq model with an encoder-decoder RNN. The chatbot leverages NLP techniques such as tokenization, word embedding, and attention mechanisms for improved response generation. The model was trained on conversational data to facilitate more natural and context-aware interactions.
🔗 Read the full report | GitHub Link

Content & Text-Based Image Retrieval and ERP Design Automation

📅 November 2018
Designed a system that retrieves images based on text descriptions using CNN and LSTM models. It also featured an ERP system for MNNIT, including seminar hall management, guest house management, and faculty intellectual property data management. The retrieval system used BLEU scores for similarity comparison.
🔗 Read the full report | GitHub Link

Waste Classification System using Convolutional Neural Networks

📅 April 2018
Developed an automated waste segregation system using Convolutional Neural Networks (CNN). The system classifies waste into five categories: container, organic waste, metal, paper, and plastic. This project aimed to improve recycling efficiency and waste management by eliminating human intervention in waste sorting.
🔗 Read the full report | GitHub Link