I am a Ph.D. Candidate in Machine Learning at Georgia Tech, advised by Prof. Juba Ziani
I'm passionate about large-scale sequential models, in particular generative retrieval and ranking systems. I also believe in safeguarding user privacy and improving performance across user segments, leveraging tools from differential privacy and algorithmic fairness.
Check out my publications and writings for more.
During my PhD, I've interned as a ML Engineer at Pinterest, an Applied Scientist at Amazon and as a MLE at Keysight.
Recent News
- 🎉 Hot off the press Check out our paper, GLoSS: Generative Language Models with Semantic Search for Sequential Recommendation, I will be presenting this at the OARS workshop@KDD2025!
- Thrilled to be joining Pinterest as a MLE intern in the Ads Candidate Generation team!
- Feb 2025: Our paper, Producers Equilibria and Dynamics in Engagement-Driven Recommender Systems has been accepted at TMLR 2025!
- Jan 2025: Our work, Improving Minimax Group Fairness in Sequential Recommendation is accepted at ECIR 2025! This is work I led during my internship at Amazon.
Experience
- Machine Learning Intern, Pinterest — Ads Retrieval (Summer 2025)
- Applied Scientist Intern, Amazon — Shopping Personalization (Fall 2024)
- Machine Learning Intern, Keysight — ML Explainability (Summer 2023)
- Research Intern, LIG — Fairness in Ad Auctions (Spring 2020)
Education
- Ph.D. in Machine Learning, Georgia Institute of Technology
Sep 2021 - April 2026
- Master's in Computer Science, École Normale Supérieure de Lyon, France
Sep 2019 - Aug 2020
- B.E. in Computer Science, BITS-Goa, India
Aug 2015 - Dec 2018
Awards
- Kiplinger Fellowship (Fall 2021, Spring 2022)
- University of Lyon Scholarship of Excellence (Fall 2019, Spring 2020)
- Merit Scholar, BITS Pilani (2015, 2016)
Teaching
- TA, ISyE 4803: Online Learning and Decision Making (Spring 2024)
- TA, ISyE 2027: Probability with Applications (Spring 2022)
- TA, ISyE 3133: Engineering Optimization (Fall 2021)
Academic Service
Reviewer: AAAI, ICML, ICLR, KDD, NeurIPS.