Ajith Suresh
Making computation private by design — cryptography researcher at the intersection of MPC, privacy-preserving ML, and federated learning.
🏛 Lead MPC Researcher · TII Abu Dhabi
📍 Abu Dhabi, UAE
🎓 Google PhD Fellow 2019
Making computation private by design — cryptography researcher at the intersection of MPC, privacy-preserving ML, and federated learning.
Three interconnected pillars that drive my work in applied cryptography
Designing efficient MPC protocols that allow multiple parties to jointly compute on private inputs — without any party revealing their data. Focus on practical, high-throughput constructions for real-world deployment.
Building MPC-based frameworks for private inference and training of neural networks. Enabling ML models to be evaluated on sensitive data without exposing the model or the data to any party.
Combining MPC with federated learning to achieve both privacy and robustness — defending against poisoning attacks and aggregation leakage in large-scale distributed ML systems.
Key contributions across MPC, PPML, and federated learning — full list with links
Tools, techniques, and concepts from a decade of cryptography research
From Kerala to Bangalore to Darmstadt to Abu Dhabi
Leading MPC research at the Cryptography Research Center (CRC), driving protocol design and applied cryptography initiatives.
Research on privacy-preserving ML for IoT, large-scale federated learning, and circuit synthesis for secure computation.
Post-doc under Prof. Thomas Schneider. Produced FLUTE, Don't Eject the Impostor, WW-FL, SafeFL, and PrivMail.
Thesis on MPC for small populations with applications to PPML, under Prof. Arpita Patra. Google PhD Fellow 2019. Outstanding thesis commendation.
Research master's in Secure Multi-Party Computation, laying the groundwork for ASTRA and early PPML protocols.
Best Outgoing Student in CSE (P Rathnaswamy Memorial Endowment). AIR 807 in GATE 2014.
Open to invited talks, collaboration opportunities, in-depth technical discussions, and exploring joint research. Drop an email or connect on social media.