Academic Coursework & Laboratory Prototypes
Semester Projects.
A comprehensive repository of cybersecurity systems, cryptographic protocols, Web3 architectures, and forensic utilities engineered throughout my coursework at FAST-NUCES Islamabad.
Secure Online Voting System
CY-4001 Secure Software Design
“A military-grade secure electronic voting platform with immutable audit trails.”
Engineered a secure Flask-based online voting system implementing end-to-end client-side encryption using hybrid RSA-4096 and AES-256-GCM. Designed an immutable, blockchain-inspired audit log chain secured with HMAC-SHA256, protecting against database-level tampering. Developed a rigorous 127-case automated test suite validating OWASP Top 10 mitigations, RBAC decorators, Argon2id password hashing, and TOTP 2FA.
Metrics & Outcomes
100% pass on 127 automated security/endpoint tests, completed comprehensive STRIDE/PASTA threat models.
DE-PWNED: Decentralized Breach Registry
Blockchain and Cryptocurrency
“A privacy-preserving, censorship-resistant Web3 alternative to Have I Been Pwned.”
Developed a decentralized, tamper-proof credential breach notification registry on the Ethereum blockchain to eliminate centralized single points of failure. Implemented custom Solidity smart contracts managing an immutable breach ledger without exposing individual credential data. Integrated MetaMask and Ganache to facilitate zero-trust, privacy-preserving credential breach verification for end-users.
Metrics & Outcomes
Deployed and verified smart contracts on a local Ganache testnet, enabling secure, real-time breach lookups with zero exposure of raw credentials.
AI-Augmented Malware Classification
Vulnerability Assessment & Reverse Engineering
“Automating malware family classification using static/dynamic reverse engineering and machine learning.”
Built a pipeline that extracts static (opcode n-grams, PE headers, API imports) and dynamic (registry writes, processes, network connections) features from malware binaries using Ghidra and sandbox logs. Manually engineered feature vectors to train machine learning classifiers capable of identifying malware families. Integrated Sentence-BERT to embed sandbox description logs, clustering malicious behavior patterns automatically.
Metrics & Outcomes
Evaluated classification performance on VirusShare malware logs, achieving high precision/recall across multiple families.
Edge Databases in IoT Security
CY-3002 Database Systems
“An applied hybrid-encrypted SQLite/MongoDB database prototype with real-time ML anomaly detection.”
Proposed and prototyped an optimized three-tier security architecture to protect IoT edge databases from interception, tampering, and denial-of-service attacks. Developed an SQLite/MongoDB local database layer utilizing hybrid AES/RSA lightweight encryption and Attribute-Based Access Control (ABAC). Integrated an ML-based anomaly detection engine at the database layer trained and evaluated on the UNSW-NB15 intrusion dataset.
Metrics & Outcomes
Achieved a 60% reduction in query latency (25ms vs 65ms) and verified an 85% attack detection rate with a 7% false-positive rate under simulated high-load conditions.
Forensic SSD TRIM Recovery Tool
CY-3006 Digital Forensics
“Programmatic extraction and data recovery from drives subject to TRIM active states.”
Developed a customized forensic recovery tool designed to extract and reconstruct deleted file structures from SSDs and HDDs that are actively subject to TRIM commands. Bypassed standard OS sector-wiping assumptions by analyzing low-level NTFS/FAT32 filesystems, recovering orphaned metadata and Master File Table (MFT) records. Implemented cryptographic integrity checking for all recovered evidence to ensure admissibility in legal proceedings.
Metrics & Outcomes
Successfully parsed raw disk images from NIST CFREDS databases, recovering orphaned file structures and maintaining a strict, verifiable cryptographic chain of custody.
Smart City Waste Routing Optimization
CS-2001 Data Structures
“Custom C++ graph-driven routing optimizer with real-time sensor-triggered path recalculation.”
Designed and built a high-performance routing optimization engine for simulated IoT-connected smart cities. Implemented custom Graph data structures, Priority Queues (using a custom Binary Min-Heap), and Dijkstra's shortest-path algorithm entirely from scratch in C++. Programmed a real-time event system allowing edge sensors to push capacity alerts, dynamically recalculating the optimal collection path on the fly.
Metrics & Outcomes
Islamabad-based 56-vertex graph, managing 50 IoT bins, 5 dumping sites, and 5 depots with seamless dynamic recalculation.
Cybernary Platform
CS-3005 Web Engineering
“A specialized, secure online education platform bridging academia and hands-on cybersecurity.”
Architected Cybernary, a modern cybersecurity learning management platform styled similarly to Udemy but containing active, hands-on labs. Implemented a secure backend that safeguards student data, logs course progression, and sandbox-runs basic cybersecurity tools. Focused on creating an affordable, high-quality channel to translate theoretical academic cybersecurity lessons into immediate, practical skill sets.
Metrics & Outcomes
Successfully developed a functional prototype featuring secure user registration, course catalogs, interactive quizzes, and embedded practical security utilities.