🧪 Research

I work broadly in the area of Trustworthy and Privacy-Preserving Machine Learning, with a strong focus on making machine learning systems secure, fair, and usable in real-world settings. My research spans both theoretical foundations and practical deployments, and I’m always looking for curious students and motivated collaborators to join in solving open problems.

🔍 Core Research Areas

🧠 Federated Learning (FL)

• Designing adaptive FL algorithms that are secure against data and model poisoning attacks.
• Exploring vulnerabilities in collaborative learning systems and proposing solutions.
• Developing inference attack models to audit information leakage in federated setups.
• Designing verifiable schemes to ensure the integrity of aggregation results.
• Optimising computational and communicational cost in real-time deployments.
• Study in cross-silo (large institutions) and cross-device (edge devices) settings.

🔐 Privacy-Preserving Machine Learning

• Leveraging Differential Privacy, Homomorphic Encryption and Secure Multi-party Computation to train models on sensitive data without revealing inputs.
• Employing Zero-knowledge Proofs to ensure integrity on computations.

✅ Trustworthy AI

• Investigating bias, fairness, and interpretability of AI systems.
• Creating tools to detect overfitting, distributional shifts, and model leakage.

🔄 Private Data Sharing & Governance

• Developing practical data-sharing systems across international boundaries, ensuring regulatory compliance.

🎓 Openings for Students

If you’re a student [undergraduate, master’s, or PhD aspirant] interested in working with me and making me learn, on closely aligned with any of the above topics, please feel free to email me, with a statement of purpose.

🤝 Looking for Collaborators

I’m looking forward to new collaborations, especially with:

• Enthusiasts passionate about the startup ecosystem for innovation
• Researchers working in TIPS [trust/identity/privacy/security]
• Industry partners interested in deploying privacy-aware applications
• Interdisciplinary teams combining law, policy, and technology

Let’s build safer and more inclusive AI systems together.