Final year B.Tech Computer Science student specializing in Data Science, currently working as a Data Science Intern. I build practical projects across data analysis, machine learning, deep learning, NLP and automation, and I like taking ideas all the way from raw data to a working deployed product.
I am actively looking for Data Analyst and Data Science roles . My work usually starts with a business or health related question, moves through data cleaning and analysis, and ends with a dashboard, app or report that someone can actually use.
Languages
Machine Learning and Deep Learning
Scikit learn, XGBoost, LSTM, ARIMA, Isolation Forest, K Means Clustering
NLP and AI
TF IDF, Cosine Similarity, LangChain, Whisper, Large Language Models (Gemini, Mistral)
Automation
Make, workflow agents connecting APIs and AI models
Data Analysis and BI
An end to end machine learning app that predicts customer churn using classification models, with about 91 percent recall. Includes data preprocessing, EDA, model building and a live deployment using FastAPI and Streamlit on Render.
A hybrid ARIMA and LSTM model built to forecast Nifty 50 stock prices, achieving close to 0.96 R squared by combining statistical time series methods with deep learning.
A full analytics project on a large hospital dataset with over 100,000 records. Covers Python and Pandas based cleaning, SQL analysis in MySQL, and an interactive Power BI dashboard, packaged with documentation and a presentation.
A set of data science tasks completed during my internship, including hypothesis testing on national survey data, patient clustering using K Means, anomaly detection in prescribing data using Isolation Forest, and an ETL pipeline processing over 770,000 records from NHS England.
A content based recommender using TF IDF and cosine similarity to suggest similar movies, served through a FastAPI backend.
A multi agent automation workflow built with n8n, connecting YouTube and SerpAPI data sources to Gemini for content generation, with results logged into Google Sheets.
A tool that transcribes and summarizes meetings using yt dlp and Whisper for audio processing, combined with LangChain, Mistral and ChromaDB for question answering, wrapped in a Streamlit interface.
An analysis of inventory performance to identify overstock exposure, measure holding cost reduction and validate service level efficiency, using SQL for business logic and Power BI for reporting.
Completing my final year of B.Tech in Computer Science with a Data Science specialization, working as a Data Science Intern, and applying to Data Analyst and Data Science roles.
