Ashrya Agrawal

Ashrya Agrawal

Data Science Intern

JP Morgan Chase & Co.

BITS Pilani - Pilani Campus

About me

Ashrya Agrawal is a Data Science Intern at JP Morgan. He is also a undergraduate Computer Science student at BITS Pilani, Pilani Campus. He is passionate about mitigating bias in Machine Learning models. His journey of fairness began with JSoC 2020 where he developed a toolkit for bias audit and mitigation.

Research in Fairness. While working on the debiasing algorithms in the JSoC toolkit, he realized that the debiasign algorithms often fail in practice to generalize out-of-sample. This piqued his interest, and he was curious to understand these failures. In his research he demonstrated that the failure of debiasers can be explained as a consequence of large variance due to the bias-variance trade-off. To mitigate this large variance and improve generalization, he formulated the notion of partial debiasing.

Research in Causality. After studying debiasing from an observational viewpoint, he studied it from a causal perspective to analyze the behavior of debiasing methods and draw deeper insights than what is provided by statistical analysis of fairness. He studied the effect of debiasing methods on causal path-specific fairness metrics. Then he explained these effects from an Information-theoretic perspective using partial information decomposition.

Apart from fairness and causality, he is interested in exploring promising research areas in Machine Learning. He has worked in ADAPT Lab on Video Captioning using knowledge bases. Ashrya is also passionate about ML reading groups, open collaborations and teaching.


  • Algorithmic Fairness
  • Causal Inference
  • Generalization of ML models
  • Video Captioning


  • B.E. Computer Science, 2022

    BITS Pilani - Pilani Campus

  • Minor - Data Science, 2022

    BITS Pilani - Pilani Campus



Data Science Intern

JP Morgan Chase & Co.

Jan 2022 – Present Bangalore
  • Working with the Corporate Investment Banking(CIB) R&A Data science team to optimize the corporate workflows using Data Science and NLP.

ML Intern

Wipro Ltd.

May 2020 – Jun 2020 Bangalore
  • Worked on automatic FAQ-answering system to improve response time for queries by employees.
  • Fine-tuned BERT and computed feature representation using the fine-tuned BERT model

Research Collaboration

with Dr. Jiahao Chen and Prof. Sebastian Vollmer

May 2020 – Dec 2021 Remote
  • Used causality to identify failure modes of debiasing algorithms and explained those failures using Information theory.
  • Benchmarked fairness algorithms on 9 datasets and demonstrated practical considerations like generalization error.
  • Worked on proof of impossibility results using Fairness-Tensor framework.
  • Performed mathematical analysis of bias-variance-noise decomposition for naïve post-processing to supplement empirical results.
  • Developed Fairness.jl, a comprehensive toolkit in Julia for bias audit and mitigation, funded by JuliaComputing.

Research Assistant

Advanced Data Analytics and Parallel Technologies Lab, BITS Pilani

Jan 2020 – May 2021 Pilani
  • Developed encoder and decoder architectures for knowledge insertion in Video Captioning models.
  • Demonstrated higher semantic consistency and improved CIDEr by 5.77 on MSVD and by 0.63 on MSRVTT over the baseline.
  • Used GCNs and various Spatio-Temporal and Reinforcement Learning methods for performance improvements.
  • Leveraged advancements in attention architectures for knowledge selection.

Teaching Assistant

Computer Science Department BITS Pilani

Aug 2019 – Nov 2019 Pilani
  • Conducted 10+ tutorial sessions with 40+ freshmen, introducing them to C programming & assisted in grading and conducting exams.
  • Taught practical aspects of Data Structures & Algorithms and conducted 10+ lab sessions for 50+ students.


IBM HackChallenge National Winner

Created Friend Affinity Finder which uses social media data to perform NLP analysis.

  • Using social media data, it gives insights about a friend’s Big 5 personality traits, general tone and interests.
  • This information is then used by clustering module to compute friendship affinity and subsequently categorize friends.
See certificate

JEE Mains

Secured 99.94 percentile (All India Rank 576) while competing with 1 million students.

KVPY Fellow SA & SX

Recipient of the prestigious KVPY fellowship twice.

  • All India Rank 185 while competing with about 60,000 other students in grade 12.
  • All India Rank 685 while competing with 50,000 other students, in grade 11.

Top 1% : National Standard Examination in Physics

Top 1% in the qualifying examination for International Physics Olympiad.
See certificate

Top 1% : National Standard Examination in Chemistry

Qualified to next stage for Indian National Chemistry Olympiad while securing top 1% marks in the qualifying exam for International Chemistry Olympiad.
See certificate



EShopping Collab

We built a web app which shows you the list of various people around you who are willing to combine the orders for e-shopping in order to save delivery charges.

Voice Assistant for Alzheimer

Developed a voice assistant with automatic Object Detection to aid navigation, memory issues for Alzheimer patients using ResNet, cloud, NodeJS, etc. Helps patients by locating objects misplaced or forgotten by patient.

FAQ Answering using BERT

Developed an automatic FAQ-answering system to improve response time for queries by employees. Fine-tuned BERT and computed feature representation using the fine-tuned BERT model

JagScript Compiler

Built a compiler in C, supporting various data-types, arrays, expressions and specifically jagged arrays. Implemented lexical analyser, parser, abstract syntax tree generator and type checker.

Checkmate Backend

Developed Backend for Checkmate 2019, a bi-annual puzzle-based event by BITS ACM.

Video Caption Generation

Video captioning is a challenging task of modelling the objects, their temporal information and interaction in order to generate a textual description. Current models often fail to model these objects and their interactions correctly, due to lack of knowledge about them.

Benchmarking Debiasing algorithms

We present an empirical study of debiasing methods for classifiers, showing that debiasers often fail in practice to generalize out-of-sample, and can in fact make fairness worse rather than better.

Causal analysis of debiasers

We study the effects of different debiasing methods on the underlying causal path specific effects (PSEs) and explain these effects using an information-theoretic perspective.

Fairness toolkit

We present Fairness.jl, a comprehensive bias audit and mitigation toolkit in julia. Extensive support and functionality provided by MLJ.jl has been used in this package.

Friend Affinity Finder

Created a Friend Affinity Finder that performs NLP analysis of the social media profiles of friends, generates affinity score for friends and clusters them.






Backend Development

Data Science