Nitin Agrawal

Nitin Agrawal is a specialist in privacy-preserving machine learning, currently working as a Privacy Engineer at Snap Inc. His work focuses on developing privacy verification and risk measurement frameworks to fortify machine learning systems internally. This includes identifying vulnerabilities, auditing models and data pipelines, and designing proactive mitigations to reduce user risk. He has extensive experience with privacy-enhancing technologies such as cryptographic methods and federated learning, while ensuring that AI systems remain functional and align with long-term regulatory, societal, and business objectives. He earned his PhD in Computer Science from the University of Oxford, where he explored the design and implementation of privacy-preserving and fair AI systems. He is dedicated to advancing responsible AI development by embedding privacy, security, and fairness principles into these systems.

• 2023–Present·Snap Inc.·Privacy Engineer
  • Responsible for privacy verification, ensuring security of machine learning systems through vulnerability detection and proactive privacy mitigations. Focused on privacy risk measurement and auditing, while enhancing fairness, safety, and trust in AI systems.
• 2022–2023·Amazon·Applied Scientist II (Privacy & Machine Learning)
  • Audited and quantified privacy leaks in machine learning systems and designed privacy-preserving machine learning solutions for audio, vision, and language models.
• 2021–2022·University of Oxford·Research Fellow
  • Conducted research on decentralized privacy-preserving computation using multi-party computation and federated learning.
• 2021·Callsign·Consultant (AI Privacy & Authentication)
  • Provided technical consultancy in privacy and authentication domains.
• 2020·Amazon Alexa AI·Privacy-Preserving AI Research Intern
  • Developed privacy-preserving on-device AI technologies.
• 2019·Microsoft Research·Privacy-Preserving AI Research Intern
  • Conducted research on privacy-preserving AI technologies.
• 2018·Alan Turing Institute·Privacy-Preserving Machine Learning Research Intern
  • Focused on research in privacy-preserving machine learning.

• 2017–2021·University of Oxford·DPhil (PhD)
  • Research Focus: Efficient and Fair Privacy-Preserving Machine Learning
  • Supervisor: Sir Nigel Shadbolt
• 2016–2017·University of Oxford·Master's
  • Research Focus: Synthesizing Realistic Quantified Self Data Using Deep Neural Networks
• 2012–2016·University of Delhi·Bachelor's (Innovative Information Technology and Mathematics)