Machine Learning, Data Science, Business Analtics

SHAURYA GULATI

412-***-**** *************@***.*** LinkedIn/shauryagulati

EDUCATION

Carnegie Mellon University Pittsburgh, Pennsylvania August 2024 - December 2025 Master's in Information Systems Management (Business Intelligence and Data Analytics) CGPA: 3.49/4.0

● Coursework: Computational Data Science, Machine Learning, Operationalizing AI

● Data Analyst at Carnegie Mellon Libraries for the Research Information Management System Project Chandigarh University Mohali, India July 2018 - June 2022 Bachelor's of Engineering in Computer Science and Engineering CGPA: 3.4/4.0 Specialization in Artificial Intelligence and Machine Learning SKILLS

Languages & Tools: Python, SQL, Git, Bash, Jupyter, VSCode, Cursor Libraries & Frameworks: PyTorch, scikit-learn, Pandas, NumPy, OpenCV, Transformers, Matplotlib, Seaborn, NLTK, BERT Data & ML Engineering: MLflow, Evidently, Azure Services, Docker, MongoDB, Tableau ML Techniques: Feature Engineering, Entity Resolution, A/B Testing, Data Visualization, Model Evaluation, Model Monitoring WORK EXPERIENCE

YMGrad June 2022- July 2024

Software Development Engineer

● Engineered a robust API in Express and Node.js, enabling seamless data retrieval from a MySQL database, resulting in a 25% faster front-end loading time and enhanced user experience

● Utilized techniques such as indexing optimization, query refactoring, and restructuring joins for improved performance, leading to a 30% reduction in data fetching time

● Conducted data analysis on user queries and platform usage trends to support internal operations teams in optimizing customer support workflows and content strategy

The Sparks Foundation January 2021- March 2021

Data and Business Analyst- Intern

● Optimized a machine learning-based credit card fraud detection model using Python and SQL, analyzing fraud patterns to identify risks; led a 3-member team to automate detection with Python scripts, increasing detection speed by 30%

● Designed and implemented fraud prevention measures using anomaly detection, reducing fraud vulnerability by 20% PROJECTS

BERT-based Question Answering System

● Fine-tuned a BERT-base model on the SQuAD v1.1 dataset to build a context-aware question answering system, achieving a 60% F1 score on the evaluation set

● Designed a modular pipeline for preprocessing, training, and inference using Hugging Face Transformers and PyTorch, enabling reproducible experimentation

● Deployed a fine-tuned BERT question answering model as a REST API using Azure ML and Azure Container Instances (ACI), enabling real-time inference

UCI Air Quality Prediction System

Designed and implemented a simulated real-time air quality prediction pipeline using the UCI Air Quality dataset, forecasting pollutant concentrations from sensor readings

● Integrated Kafka for streaming data ingestion and built a robust training workflow with MLFlow for experiment tracking and reproducibility. Trained models, including XGBoost, achieved an accuracy of 91% in the predictions

● Containerized the prediction service using Docker, and deployed an API for serving real-time predictions

● Developed a monitoring dashboard with Evidently AI to track data drift and model performance in production Mini RAG System for YouTube Summarization

● Developed a lightweight Retrieval-Augmented Generation (RAG) pipeline to summarize YouTube videos using transcript data and Google Gemini Pro.

● Leveraged YouTube Transcript API for document retrieval and applied prompt-based summarization to simulate the RAG workflow with external data.

● Built an interactive Streamlit frontend with transcript extraction, thumbnail preview, and summary generation in real-time. PUBLICATIONS

● "Enhancing Sentiment Analysis in Short Texts with POS-Embedded LSTM Models," IATMSI 2024

● "Paint/Writing Application through Webcam using MediaPipe and OpenCV," ICIPTM 2022

Contact this candidate