Helped rewrite the ‘Customer Setup’ platform, modernizing the old IBM AS/400 mainframes using AngularJS
Seamlessly incorporated over 30 essential core banking modules - including ChexAdvisor, FraudChex
Customer Search, and Mergers/Acquisitions/Splits - enhancing the application’s functionality and efficiency
Reduced application access time by 60% through the implementation of GUI interfaces, reports & dashboards
Engaged in cross-functional team collaboration, participated in product discussions and code reviews, and played a key role in both acceptance testing and go-live production builds.
Summer Research Intern
Bennett University
05.2020 - 06.2020
Edge Computing, Custom k-means implementation
Completed a 2-month virtual internship in Deep Learning and Artificial Intelligence
Collaborated on feature pruning techniques that removed ∼40-50% of parameters, effectively optimizing Neural Networks for Mobile/Edge devices using PyTorch
Attained a 30% FLOP reduction on CIFAR-100 dataset, without affecting the accuracy of the VGG-16 model
Education
Master of Science - Data Science
Indiana University
Bloomington, IN
08.2023 - Current
B.Tech - Computer Science and Engineering
Siddharth Institute of Engineering & Technology
07.2021
Skills
Languages: Python, R, SQL, Java, JavaScript, C
Data Science: EDA, Feature Selection & Engineering (Numpy, Pandas), Data Visualization (Matplotlib, Tableau)
Data Modelling, Machine Learning (CNN, Linear & Logistic Regression, Naive Bayes, KNN)
NLP: Open-source LLMs, Info Retrieval (RAG), Fine-tuning, ChatBots, LangChain, Ollama
Web Tech: Angular, React, Flask, Nodejs, RESTful API, HTML, CSS
Dev Tools: Git/BitBucket, PyTorch, AWS, Azure, GCP, Docker, CI/CD, Postman, Jira, MS Office Suite
Accomplishments
ChatGPT and LangChain: The Complete Developer’s Masterclass | Udemy
Natural Language Processing - Specialization | Coursera
TensorFlow in Practice - Specialization | Coursera
Microsoft Certified: Azure Fundamentals | Microsoft
Agile Development Practices | LinkedIn Learning
Regulatory and Compliance Trainings | FIS Global
Publications
Detection of Online Toxic Comments Using Deep Learning Preprint | Study paper, ICRDBI-2021
Presented a study paper at the ICRDBI conference examining the capabilities and limitations of various
NLP-related architectures, including Probabilistic (Na¨ıve Bayes), Sequential (RNN, LSTM, GRU), and Attention
(Transformers) models, for their efficacy in toxicity classification.