Hello, I’m Shashank.
I’m a Senior Applied Scientist at Microsoft AI. I build machine learning systems that learn safely and effectively from human behavior.
A little about me.
I am a Senior Applied Scientist at Microsoft AI, working on machine learning for ads. My broader research asks a practical question: how can intelligent systems learn safely and effectively from the people who use them?
I completed my PhD at the Information Retrieval Lab at the University of Amsterdam, supervised by Prof. Maarten de Rijke and Prof. Harrie Oosterhuis. My thesis focused on off-policy evaluation and learning for ranking and recommendation systems.
During my PhD, I worked with Meta AI’s Modern Recommender Systems teams in London and New York on two-stage recommendation and reinforcement learning for text-to-image diffusion models. Before that, I built search-ranking and query-understanding systems at Flipkart.
Experience
Microsoft AI
Senior Applied Scientist
Ads · Bengaluru
Meta AI · New York
Research Scientist Intern
RL for diffusion models
Meta AI · London
Research Scientist Intern
Two-stage recommendation
Flipkart
Data Scientist
Search ranking & query understanding
Conduent Labs (erstwhile Xerox Research)
Research Intern
Fake-news detection
Tata Research Design and Development Center (TRDDC)
Research Intern
Adverse-drug reaction extraction
University of Amsterdam
PhD · Information Retrieval
Ranking & reinforcement learning
IIIT Hyderabad
MS by Research
Search & information extraction
Latest updates
Research, talks, and milestones
2026
- [June] Our SIGIR 2026 paper on off-policy evaluation was featured on Microsoft Research’s official newsletter, Microsoft Research Newsletter.
- [April] Joined Microsoft AI, Ads team in Bangalore as a senior applied scientist.
- [April] One short paper on the theoretical analysis of our previously proposed $\beta$-IPS (from the RecSys 2024 paper) accepted at SIGIR 2026.
- [March] Paper from my Meta AI internship on A Simple and Effective Reinforcement Learning Method for Text-to-Image Diffusion Fine-tuning was accepted at Transactions on Machine Learning Research (TMLR).
- [Feb] PhD dissertation abstract published in the December 2025 edition of SIGIR Forum.
- [Jan] Wrapped up a paper where we showed that $\beta$-IPS (a novel estimator we proposed in our RecSys 2024 paper) is asymptotically better than the SNIPS estimator in terms of MSE. preprint.
- [Jan] Presented my work on RL for recsys and post-training diffusion models at eBay, San Jose, virtually. slides.
2025
- [Oct] Successfully defended my PhD thesis on reinforcement learning for ranking and generative models. The video recording of the defense presentation is available here. P.S.: I have not gone through the full recording myself — it’s painful to listen to my own voice :)
- [Aug] Invited to talk about my work on RL for recommendation and diffusion models at LossFunk, Bangalore. The slides are available here.
- [June] Research paper from my first internship at Meta, “Towards Two-Stage Counterfactual Learning to Rank”, was accepted at ICTIR 2025 (co-located with SIGIR).
- [March] The preprint of my work from the internship at Meta AI, NYC, “A Simple and Effective Reinforcement Learning Method for Text-to-Image Diffusion Fine-tuning” is now available.
Earlier updates 2017—2024
2024
- [October 2024] Presented my work on Safe Deployment for Counterfactual Learning-to-Rank at Expedia, London (virtually).
- [September 2024] Two papers “A Simpler Alternative to Variational Regularized Counterfactual Risk Minimization”, and “Proximal Ranking Policy Optimization for Practical Safety in Counterfactual Learning to Rank” were accepted at the CONSEQUENCES ‘24 workshop @RecSys.
- [September 2024] We released the tutorial recording on Recent Advancements in Unbiased Learning to Rank, previously presented at WSDM 2024, SIGIR 2023, and FIRE 2023.
- [August 2024] Joined Meta AI, New York as a research scientist intern for the summer. I’ll be working on reinforcement learning-based fine-tuning for text-to-image diffusion models.
- [July 2024] Full paper on “Optimal Baseline Corrections for Off-policy Contextual Bandits” was accepted at RecSys 2024, with an oral presentation.
- [July 2024] Full paper on “Practical and Robust Safety Guarantees for Advanced Counterfactual Learning to Rank” was accepted at CIKM 2024.
- [May 2024] PC member for ICML 2024, ICLR 2024, SIGIR 2024, RecSys 2024, and ICTIR 2024.
- [May 2024] Arxiv preprint of our work on “Optimal Baseline Corrections for Off-policy Contextual Bandits” is available online here.
- [March 2024] Slide deck from the WSDM tutorial on recent advancements in unbiased LTR is available here.
- [March 2024] I will be co-presenting a tutorial on recent advancements in unbiased LTR at the WSDM 2024 conference.
2023
- [Oct 2023] Tutorial proposal on recent advancements in unbiased learning-to-rank accepted at WSDM 2024, and FIRE 2023.
- [Sept 2023] Invited talk at ShareChat on Safe Unbiased Learning-to-Rank slides, video.
- [August 2023] Invited talk at Meta AI, New York on Safe Unbiased Learning-to-Rank.
- [July 2023] Joined Meta AI, London as a Research Scientist Intern, where I’ll be working with the Modern Recommender Systems (MRS) team on off-policy learning for two-stage recommender systems.
- [July 2023] Two papers (extended abstracts) were accepted at the CONSEQUENCES workshop, co-located with RecSys’23. First work is a SIGIR resubmission on safe unbiased learning-to-rank; second work is on examining & mitigating selection bias in preference elicitation for recommender systems.
- [June 2023] Paper on “A Deep Generative Recommendation Method for Unbiased Learning from Implicit Feedback” was accepted at ICTIR’23, co-located with SIGIR’23.
- [May 2023] Paper on user return time prediction was accepted at the CRUM workshop at UMAP 2023. This paper was part of the work done while I was at Flipkart, India.
- [April 2023] Tutorial proposal on recent advancements in unbiased learning-to-rank was accepted at SIGIR’23. I will be leading the tutorial discussion at the conference.
- [April 2023] One full paper on safe unbiased learning-to-rank was accepted at SIGIR 2023.
- [Feb 2023] PC member for SIGIR’23, NeurIPS’23.
2021-2022
- [August 2022] Work on VAE-IPS was accepted at the CONSEQUENCES+REVEAL Workshop (as oral presentation) at RecSys’22.
- [June’21 - June’22] Reviewed for ACL, EMNLP, ICLR, ICML, NeuRIPS’22.
- [Sept’21 - Nov’21] TA for the Advanced Information Retrieval course.
- [May 2021] Reviewing for NeurIPS, EMNLP, ICLR 21.
- [April 2021] Joined IRLab, UvA as a PhD student
Pre-PhD
- [July 2020] PC Member for EACL’2020.
- [April 2020] Position paper accepted at the ECNLP@ACL’20.
- [April 2020] Paper accepted at SIGIR’20.
- [April 2020] PC Member for EMNLP’20 (IE Track) & AACL-IJCNLP 2020 (IE + IR Track).
- [Dec 2019] PC Member for WSDM’20 Workshop on State-based User Modelling.
- [Nov 2019] Reviewing for ICML 2020. [Update as of March ‘20] - Couldn’t review papers due to bad health.
- [July 2019] Program Committee Member for ECIR 2020.
- [June 2019] Reviewing for a special edition “Learning from User Interaction” of Information Retrieval Journal (IRJ).
- [March 2019] Invited talk at Alumni Research Talks event organized at BITS, Pilani. I talked about “Information Retrieval from Social Media” slides.
- [Oct 2018] Serving as PC member for ECIR’19 and ML4H@NeurIPS’18.
- [July 2018] Joined Flipkart as a Data Scientist with the Search team.
- [June 2018] Internship work @Conduent Labs on “Fake News Detection” was accepted at ASONAM’18 (Short Paper).
- [June 2018] Journal paper from my summer internship work @TRDDC, Pune was published at BMC Bioinformatics (Impact Factor: 2.448).
- [June 2018] Serving as PC Member for ICON 2018 and DTMBio 2018 @CIKM’18.
- [January 2018] Joined Conduent Labs, Bangalore (previously known as Xerox Research Center India (XRCI)) as Research Intern with Manjira Sinha and Sandya Mannarswamy.
- [December 2017] Two papers were accepted at ECIR 2018.
- [November 2017] Two papers were accepted at the NeurIPS 2017 Machine Learning for Health (ML4H) Workshop.
- [November 2017] Invited talk on “Deep Learning for Recommender Systems” at Thiagarajar College of Engineering (TCE) and “Machine Learning & Information Retrieval Techniques for Adverse Drug Reaction Mention Extraction from Social Media” at Duke-NUS Medical School, Singapore.
- [September 2017] Two papers on Content-Based News Recommendation Systems were accepted at the ICDM workshop on Semantic Recommendation Systems (SeRECSys).
- [August 2017] Paper on ‘Semi-Supervised Recurrent Neural Network for Adverse Drug Reaction Mention Extraction’ was accepted at ACM 11th International Workshop on Data and Text Mining in Biomedical Informatics at CIKM 2017.
- [July 2017] Paper on Knowledge Base integration with text classification pipeline was accepted at the SIGIR workshop (KG4IR).
- [June 2017] Paper on Trust Prediction in Social Media using Neural Networks was accepted at ASONAM’17.
- [May 2017] My work on Hate Speech Detection from social media was covered in some leading publication houses in India. Source
- [April 2017] I received the Best Poster award at WWW’17 for my work on Hate Speech Detection from social media.
- [Feb 2017] Poster paper accepted at WWW’17.
- [Feb 2017] Workshop paper accepted at WWW’17.