Research Scientist / Engineer
1 Hacker Way
Menlo Park, CA, 94205
email: ananthr (at) cs (dot) stanford (dot)
I am a computer scientist broadly interested in cryptography, privacy, and their intersections with Machine Learning. I work on Privacy Infrastructure at Facebook.
Before Facebook, I spent nearly six years at Google working on privacy and machine learning research as part of Google Brain. Some of the projects I worked on or helped with include local differential privacy, the the Shuffle Model of privacy and its privacy amplification guarantees, the Chrome Privacy Sandbox proposals, algorithms to privately train ML models, Chrome's Password Check tool, and post-quantum crypto (as part of the FrodoKEM team that designed and submitted the Frodo Key Encapsulation Mechanism to the NIST competition on post-quantum crypto).
I received my Ph.D. from the Department of Computer Science at Stanford University advised by Prof. Dan Boneh. My thesis focused on modeling and building secure deterministic and searchable encryption schemes. I also worked on building lattice-based cryptographic primitives, among other topics in cryptography.
During my Ph.D., I spent summers with Dirk Balfanz and the security engineering team doing research on the Security Key at Google, and at Microsoft Research (Silicon Valley) working with Gil Segev and Ilya Mironov. Earlier, I graduated from the Computer Science and Engineering Department at IIT Madras with a Bachelors of Technology in Computer Science.
Protecting accounts from credential stuffing with password
With Kurt Thomas, Jennifer Pullman, Kevin Yeo, Patrick Gage Kelley, Luca Invernizzi, Borbala Benko,
Sarvar Patel, Dan Boneh, and Elie Burzstein
Distinguished Paper Award Winner
USENIX Security 2019
Amplification by Shuffling: From Local to Central Differential Privacy via Anonymity
With Úlfar Erlingsson, Vitaly Feldman, Ilya Mironov, Kunal Talwar, and Abhradeep Guha Thakurta
Scalable Private Learning with PATE |
With Nicolas Papernot, Shuang Song, Ilya Mironov, Kunal Talwar, and Úlfar Erlingsson
Prochlo: Strong Privacy for Analytics in the Crowd |
With Andrea Bittau, Úlfar Erlingsson, Petros Maniatis, Ilya Mironov, David Lie, Mitch Rudominer, Ushasree Kode,
Julien Tinnes, and Bernhard Seefeld
ACM SOSP 2017
Frodo: Take off the ring! Practical, Quantum-Secure Key Exchange from
With Joppe Bos, Craig Costello, Léo Ducas, Ilya Mironov, Michael Naehrig, Valeria Nikolaenko, and Douglas Stebila
ACM CCS 2016
Information Leakage in Embedding Models
With Congzheng Song
Encode, Shuffle, Analyze Privacy Revisited: Formalizations and Empirical Evaluation
With Úlfar Erlingsson, Vitaly Feldman, Ilya Mironov, Shuang Song, Kunal Talwar, and Abhradeep Guha Thakurta
That Which We Call Private
With Úlfar Erlingsson, Ilya Mironov, and Shuang Song
Poster at USENIX Security 2019
Improved Constructions of PRFs Secure Against Related-Key
With Kevin Lewi and Hart Montgomery
Function-Private Subspace-Membership Encryption and Its
With Dan Boneh and Gil Segev
Function-Private Identity-Based Encryption: Hiding the Function in
With Dan Boneh and Gil Segev
Message-Locked Encryption for Lock-Dependent Messages
With Martín Abadi, Dan Boneh, Ilya Mironov, and Gil Segev
Key-Homomorphic PRFs and Their Applications
With Dan Boneh, Kevin Lewi, and Hart Montgomery
Deterministic Public-Key Encryption for Adaptively Chosen Plaintext
With Gil Segev and Salil Vadhan
Journal of Cryptology 2018
Algebraic PRFs with Improved Efficiency from the Augmented
With Dan Boneh, Hart Montgomery
ACM CCS 2010
Obfuscating Straight Line Arithmetic Programs |
With Srivatsan Narayanan, Ramarathnam Venkatesan
DRM Workshop at ACM CCS 2009
In the past, I have reviewed papers for the following conferences:
CRYPTO, EUROCRYPT, IEEE S&P (Oakland), Usenix Security, PKC, TCC, PODS, ICALP, Financial Crypto, and ICISC.
My sister Aditi Raghunathan is
also a computer scientist.
A short write-up on the Gödel prize-winning Toda's theorem—one of my favorite results in complexity theory—as a project report for Prof. Luca Trevisan's CS254.
My Erdös number is 3: Paul Erdös → Peter Montgomery → Ramarathnam Venkatesan → me.