π§ͺ Research
My research lies at the intersection of Trustworthy AI, Privacy, and Security, with an emphasis on building machine learning systems that are robust, verifiable, and deployable in real-world settings. I am also interested in applications of AI in science and engineering.
Below are the key areas I am currently working on.
π Current Research Focus
β Trustworthy AI (Across all verticals)
- Studying robustness, privacy, and fairness in modern AI systems.
- Evaluating vulnerabilities and defenses in large models, including LLM-based pipelines.
π Fully Homomorphic Encryption (FHE)
- Benchmarking existing industrial FHE frameworks.
- Designing practical applications of FHE for secure data processing.
π‘ Post-Quantum Cryptography (PQC)
- Exploring how PQC can be integrated with machine learning and distributed systems.
- Investigating future-proof cryptographic designs for AI-driven applications.
π€ Generative AI (GenAI)
- Exploring Generative AI for scientific applications (AI for Science and Engineering).
- Studying privacy, leakage, and misuse risks in generative models.
ποΈ Computer Vision (CV)
- Working on plant leaf disease detection using vision-based models.
- Exploring lightweight and deployable CV solutions for real-world environments.
𧬠Biometrics
- Studying privacy risks, inference attacks, and secure biometric matching.
- Exploring cryptography-backed biometric data sharing and verification.
π° Finance & Secure Data Analytics
- Studying privacy-preserving analytics for financial data.
- Designing secure and compliant ML workflows for sensitive financial datasets.
π Working With Students
I am happy to work with motivated undergraduate, masterβs, and PhD students interested in working on the above problems. If you enjoy thinking deeply, experimenting rigorously, and building systems, feel free to reach out via π§ email with a short statement of purpose, provided you have already made some progress in one of the above-mentioned areas.
π€ Collaboration
I welcome collaborations with:
- Researchers in Trust, Identity, Privacy, and Security (TIPS)
- Industry partners deploying privacy-aware AI systems
- Interdisciplinary teams spanning AI, cryptography, policy, and systems
- Interdisciplinary teams working in Biomedicals, Healthcare and Finance
Letβs work together to build AI systems we can actually trust.