Sanjana Srivastava

Sanjana Srivastava

MS in Data Science Student

I am a Masters in Data Science student at the University of Virginia, scheduled to graduate in May 2025. I have a strong background in deep learning, computer vision, multimodal learning, product analytics, healthcare research, large language models, and resposible AI product development. I previously worked as a Research Scientist at University of Virginia with Dr. Donald E. Brown and Dr. Sana Syed. My work involved using data science techniques and training machine learning models for disease classification and pattern recognition in multimodal medical data. I am currently working on a capstone project with Deloitte as a Data Scientist, focusing on evaluating large language models assessment methodologies, including NLP-based approaches and LLM-as-a-judge methods, to assess LLM-generated content across multiple dimensions. Previously I have interned at BNY Mellon and ZestMoney as a Software Developer. I also hold a Master of Technology and a Bachelor of Technology in Geological Technology and Mathematics from the Indian Institute of Technology, Roorkee.

Find my CV here.

šŸ“– Education

  • 2024 - May 2025: Masters in Data Science | University of Virginia
  • 2021 - 2022: Masters in Technology | Indian Institute of Technology, Roorkee
  • 2017 - 2021: Bachelors in Technology | Indian Institute of Technology, Roorkee

šŸ’» Experience

  • Data Scientist at Deloitte (Capstone Project) | September 2024 - Ongoing
    • Comparative study of large language model evaluation frameworks with a focus on NLP vs LLM-as-a-judge metric.
    • Analyzing large language model evaluation frameworks like AWS Bedrock, GCP Vertex, and RAGAS to assess LLM-generated content.
    • Investigating six dimensions of LLM assessment, including bias, hallucination, and toxicity detection.
  • Machine Learning Researcher at University of Virginia | July 2022 - May 2024
    • Worked with a multi-disciplinary team of medical professionals and engineers to study gut functions.
    • Implemented novel Deep Learning models for disease diagnosis and quantification.
    • Leveraged Machine Learning for pattern recognition in tissue images, clinical data, and transcriptomic data.
  • Software Engineering Intern at BNY Mellon | Summer 2021
    • Implemented functional and unit testing for internal applications.
    • Developed a custom XML to CSV Parser Utility.
    • Conducted A/B testing to assess performance optimizations in internal applications.
  • Software Engineering Intern at ZestMoney | Summer 2019
    • Implemented a Payment Gateway at checkout using SpringBoot & MySQL database.
    • Developed a custom user Signup interface using Retrofit (Android) with MVVM architecture.
    • Improved code usability in a high-paced fintech industry startup environment.

šŸ“‚ Projects

Evaluating efficacy of synthetic images generated using diffusion models
Developed diffusion models to generate histology patches conditioned on nuclei locations. Validated the use of synthetic patches for improving downstream segmentation and classification tasks.
Code

Deep learning based detection and visual understanding of diseases using medical imaging data
Performed patch-based invasive ductal carcinoma (breast cancer) and gastrointestinal disease detection. Implemented convolutional neural networks (CNN) for classifying whole slide images and biomarker data. Implemented Gaussian clustering methods to identify recurring visual patterns in diseased biopsies.
Code

Humorous Image Captioning System
Implemented a self-attentive encoder-decoder framework to generate humorous captions for images indistinguishable from human generated memes.

Correlating disease gene signature with imaging data
Designed a deep learning framework to identify image features associated with functional gene clusters. Identified important gene signatures and their correlation with visual patterns in biopsies.
Code

Deep learning based semantic segmentation on brain MR images
Performed tumor segmentation using a U-Net architecture on MRIs. Used PyTorch for training model on MRIs from The Cancer Genome Atlas (TCGA) lower-grade glioma collection.
Code

Petrographic characterisation of a chondrite sample
Investigated the mineralogy and major element geochemistry of mineral phases present in the chondrite section. Performed Electron Probe Micro Analysis to obtain backscattered electron images of the sample.

Alzheimerā€™s disease analyses using patient data
Explored factors associated with Alzheimerā€™s, developed a predictive model, and conducted statistical analyses using regression models on patient chart data.
[Code] (https://github.com/5anjana/alzheimer-linear-regression)

Omniport Registration web application
Developed an Intranet application for students, professors, and staff for Channel i registration. Channel i is the portal with intranet applications of IIT Roorkee.
Website

Buy and Sell application
Developed an intranet application for buying, selling and requesting goods among campus residents. Features include categorization of goods, subscribing, filtering, and searching based on users input.
[Code] (https://github.com/IMGIITRoorkee/omniport-app-buy-and-sell)

People Search
Developed an Intranet application to manage contact information of students and faculty of IIT Roorkee. Features consists of recursive searching and content filtering based on the users input. Implemented Student Homepages, a UI template to view each studentā€™s bio in a standard format.
Code

Chicken Litte Run
A fun run and dodge game developed on HTML Canvas where the Chicken Little must run and dodge the falling sky.
[Code] (https://github.com/5anjana/Chicken-Little-Run)

šŸ“ Publications

Machine-learning-based integrativeā€“ā€˜omics analyses reveal immunologic and metabolic dysregulation in environmental enteric dysfunction
F. Zulqarnain, X. Zhao, K. Setchell, Y. Sharna, P. Fernandes, S. Srivastava, A. Shrivastava, L.Ehsan, V. Jain, S. Raghavan, C. Moskaluk, Y. Haberman, L.A. Denson, K. Mehta, N.T. Iqbal, N. Rahman, K. Sadiq, Z. Ahmad, R. Idress, J. Iqbal, S. Ahmed, A. Hotwani, F. Umrani, B. Amadi, P. Kelly, D.E. Brown, S.R. Moore, S.A. Ali, S. Syed.
iScience 2024.
Paper

Quantitative Morphometry and Machine Learning Model to Explore Duodenal and Rectal Mucosal Tissue of Children with Environmental Enteric Dysfunction
M. Khan, Z. Jamil, L. Ehsan, F. Zulqarnain, S. Srivastava, S. Siddiqui, P. Fernandes, M. Raghib, S. Sengupta, Z. Mujahid, Z. Ahmed, R. Idrees, S. Ahmed, F. Umrani, N. Iqbal, C. Moskaluk, S. Raghavan, L. Cheng, S. Moore, S.A. Ali, J. Iqbal, S. Syed.
The American Journal of Tropical Medicine and Hygiene 2023.
Paper