Education

Masters - Aug 2021 to May 2024

Specialisation: Computer Science (CSE), University of Wisconsin - Madison.
Masters Thesis: Understanding the effects of Compression on LLMs (PDF)
Advisor: Prof. Frederic Sala

I also had great chance working with Prof. Yin Li during my Masters.

Other interesting reports:

  • Trade-offs in Learning Commonsense vs Factual Knowledge with Pretrained Language Models (PDF)

  • Dataset Augmentation using Diffusion Models (PDF)

  • Presentation for CS762, Advanced Deep Learning (PDF)

  • Presentation for CS839, Foundational Models (PDF)

Bachelor Of Technology (B.Tech.) - June 2015 to May 2019

Specialisation: Electronics and Communication Engineering (ECE), NIT Calicut.
Undergraduate Thesis: Object classification and tracking for autonomous carnavigation in Indian road scenarios (PDF)
Advisor: Dr. Praveen Sankaran

Mini Project Thesis: A Blind Assistive Device (PDF)
Advisor: Dr. Dhanaraj

Other interesting reports: Gravitational waves, Artificial Intelligence coursework

Prior to that, I did my intermediate from FIITJEE and schooling from Ravindra Bharathi Public School and Kennedy in Vijayawada.

Professional Work and Internships

AI Scientist, GE HealthCare

June 2024 - Present

  • Manager: Jun Ma and VP Danica

  • Building customer-faced and internal-used chatbots for diagnosing repairs and fixed for various medical devices.

  • Also working to understand the alignment aspects of LLMs under compression.

Applied Scientist (Intern), Amazon

May 2023 - Aug 2023

  • My manager: Brijesh

  • Created an LLM-based recommendation pipeline for Amazon's inventory, which recommends event-specific products from a catalog of 25 million items.

  • Leveraged Sentence Transformers to generate embeddings, FAISS for efficient similarity search and HDBSCAN, UMAP to create a robust pipeline which is capable of rejecting false positives.

  • Captured higher seasonal demand compared to existing pipelines at Amazon both on historical and forecast demand data.

Machine Learning Scientist (Intern), Truera

May 2022 - Aug 2022

  • My manager: Ricardo Shih

  • Worked on segment level root cause analysis (RCA) for NLP models to understand how models treat different segments differently. Implemented metrics such as Wasserstein distance, Difference of Sums on Yelp dataset to identify tokens that might be responsible for Disparate Impact (DI).

  • Used Trulens, an open-source explainability framework for NLP model probing and implemented functions to probe at a specific layer. Visualized Influence Sensitivity Plots (ISPs) for downstream debugging of neural networks.

University of Wisconsin - Madison

Teaching Assistant

Fall 2021, Fall 2022, Spring 2023, Fall 2024, Spring 2024

  • Worked as a TA for CS220 Introduction to Data Science for various semesters. Common responsibilites include grading, proctoring, preparing questions/materials, holding office hours.

Research Assistant

Spring 2022

  • Worked with Prof. Kevin Eliceiri on medical image segmentation using traditional computer vision and self supervised techniques.

Junior Research fellow (JRF), CiSTUP, IISc Bangalore

Feb 2020 - Feb 2021

Worked with Dr. Tarun Rambha on developing algorithms to reduce traffic congestion for Bangalore buses.
My work is focused towards solving “Bus bunching” problem to reduce traffic congestion.

Software Developer, Microfocus

Aug 2019 - Feb 2020

Worked as a backend developer in ZENWorks Service desk (ZSD) which is primarily a ticket-raising platform. My tasks include fixing customer bugs, developing features related to automatic database up-gradation and writing basic RESTful services.

Internship, IIST Trivandrum

Dec 2018

Worked with Dr. Deepak Mishra on Computer Vision. Project involves detecting number plates from traffic video footage. (Report PDF)

I am grateful to all the faculties who taught me as their role is vital in shaping my career.

Research Interests

I broadly classify my research interests into 2 types

  1. Interests which I pursue as part of profession/career

    1. Artificial Intelligence

    2. Machine Learning, Deep Learning (Primary focus on Computer Vision)

  2. Interests which are like a hobby

    1. Biology - Neuroscience, Genetics

    2. Physics (Especially Quantum side)

    3. Economics, Finance

    4. Mathematics

I know I can't be a master in these fields, but I learn by reading various Non Fiction books. Check the Books and Blogs tab to understand more on this.