About Me
I am a 3rd year PhD student, jointly affiliated with the University of Padua and the University of Camerino in Italy, specializing in advancing the field of Blockchain and Distributed Ledger Technology. I am part of the Security and Privacy (SPRITZ) research group, under the supervision of Prof. Mauro Conti. My research focuses on enhancing the detection of critical bugs in blockchain consensus protocols by introducing innovative probabilistic state models, context-aware mechanisms, and advanced mutation techniques.
By addressing limitations in existing frameworks, my work significantly improves the accuracy, efficiency, and coverage of bug detection, enabling the discovery of hidden, semantic, and concurrency bugs. Through my contributions, I aim to establish more reliable and secure consensus systems, driving innovation in blockchain and distributed computing.
Universities & Research Group
Resume
Education
Experience
Workshops
Ongoing Research
PIN: Application-level Consensus for Blockchain-based Artificial Intelligence Frameworks
Currently exploring Proof-of-Intelligence (PIN), an innovative AI-driven application-level consensus framework designed for blockchain and federated learning. This research focuses on ensuring quality-assured AI enablers and developing the PIN-BOARD platform to advance decentralized AI assurance and enhance blockchain-based federated learning.
A Unified Framework for Blockchain Consensus Security and Advanced Fuzzing Techniques
This research introduces an integrated framework to enhance blockchain consensus protocol security through innovative state modeling and advanced fuzzing mechanisms. A decentralized, transparent, and secure platform is developed to analyze vulnerabilities in blockchain systems. The framework incorporates a hybrid probabilistic state model that combines Markov Chains and Hidden Markov Models, enabling a detailed exploration of both high- and low-frequency messages in blockchain consensus protocols. This allows the discovery of latent states and vulnerabilities overlooked by traditional approaches. Additionally, an advanced mutation mechanism is presented, employing context-aware message pooling and prioritized mutation processes. Leveraging clustering techniques such as DBSCAN, the framework optimizes seed message selection, reduces testing time, and improves the efficiency of bug detection.
Publications
Journal Publications
Conference Papers
Workshops
Contact
Email: tannishtha.devgun@studenti.unipd.it
Office: Office 7D4, Torre Archimede, Department of Mathematics, Via Trieste 63, 35121 Padova (PD) - Italy