Cybersecurity Researcher & Patent Holder
Security Compliance & GRC Strategy | Mitigating Deepfake Vulnerabilities | Risk Management
I'm a BTech Cyber Security student at Shah and Anchor Kutchhi Engineering College with a passion for developing innovative AI/ML solutions and cybersecurity systems.
As a project leader, I've successfully managed cross-functional teams to deliver award-winning solutions including TrustTrace (deepfake detection platform), LinuxAdmin (security management system), and VerifyIT (misinformation detection).
My work has been recognized globally with the Intel AI Challenge 2025 victory, and I've published 9 research papers in Scopus-indexed journals along with securing a patent in cybersecurity and AI domains.
Current Focus: Leveraging technical expertise to drive complex technology initiatives in project management roles while continuing research in AI safety and cybersecurity.
Multimedia Deepfake Detection Platform
Award-winning platform supporting image, video, and audio deepfake detection with ensemble ML models.
Results: 92% accuracy, 500+ active users
Recognition: Global Winner - Intel AI Challenge 2025
Air-Gapped Active Directory System
Enterprise security system with master-slave architecture, centralized monitoring, and automated threat detection.
Features: Port scanning detection, privilege escalation alerts, SIEM integration
Fake News Detection Platform
Multi-modal fact-checking platform using BERT transformers for text analysis, image reverse search, and URL verification.
Accuracy: 88% classification accuracy
Winner (Global Level)
Project: Deepfake Detective - Securing Social Integrity, Uncovering Faux AI Content
Winner - Delhi Police (View Certificate: Proof)
Theme: Counteracting Misinformation and Fake News
Winner - Vidyalankar Institute of Technology
Project: TrustTrace - Multimedia Deepfake Detection (Award Proof: View)
Gold Trophy - Quality Circle Forum of India
Projects: TrustTrace & LinuxAdmin
AIR 40 - Pentathon 2025
AIR 36 - DawgCTF 2025
Achieved 97% accuracy in deepfake detection using ResNet50 with image scraping feature.
Comprehensive analysis of EfficientNetB7 achieving 85% accuracy in deepfake detection.
Achieved 94% accuracy on FaceForensics and DFDC datasets using attention mechanisms and Siamese training.
Ensemble model approach for comprehensive deepfake detection across multiple architectures.
Achieved 97.28% accuracy in audio deepfake detection using SVM and Mel-Frequency Cepstral Coefficients.
Multi-faceted approach integrating MFCC-based SVM and Neural Networks for audio verification.
Comprehensive analysis of deepfake detection methodologies using autoencoders, GANs, and CNNs.
Comprehensive analysis of deepfake detection methodologies using autoencoders, GANs, and CNNs.
System for detecting potential misinformation through manipulated content across multiple media types.
Government of India Patent Office | Filed: April 10, 2024
System and method for detecting potential misinformation spread through manipulated content.
Advanced detection system for multimedia deepfake identification.
Centralized security and management system for network-connected devices.
• Enhanced cybersecurity awareness via digital safety workshops
• Reached 200+ participants across multiple events
• Conducted training sessions on emerging cyber threats
• Developed responsive and secure websites for small businesses
• Technologies: HTML, CSS, JavaScript, Shopify
• Focus on security best practices and user experience
prathamshah123247@gmail.com
+91-9699143903
I'm always interested in discussing new projects, creative ideas, or opportunities to be part of your vision. Whether you're looking for a project manager with technical expertise or need cybersecurity consultation, feel free to reach out!
Send Email