Vidhi Bhutia Former Intern @ Morgan Stanley • Ex - NIT Rourkela, SheFi, Aspire

A

Agentic AI Researcher

Jan 2026

- Developed an autonomous AI agent capable of searching academic papers, extracting insights, and generating structured research documents. - Implemented agentic workflows including planning, tool usage, reasoning, and synthesis using Gemini AI and arXiv integration. - Built automated PDF processing and LaTeX-based document generation pipeline to produce complete research papers. - Designed a Streamlit interface to enable users to generate research summaries and documents from simple prompts.

n

n8n Job Search Automation

Dec 2025 - Jan 2026 (2 months)

- Built an end-to-end automation workflow using n8n to automatically search LinkedIn jobs based on personalized filters and preferences. - Integrated Google Sheets, Google Drive, and Telegram APIs to store job listings, parse resumes, and deliver real-time job notifications. - Designed scheduled workflows to run periodically, extract structured job data, and eliminate duplicate listings using automated tracking. - Implemented resume parsing and dynamic job matching logic to improve job discovery efficiency and reduce manual search effort.

H

Health Analysis using Federated Learning and Cloud

Nov 2025 - Jan 2026 (3 months)

- Built a privacy-preserving health risk prediction system using Federated Learning, enabling multiple hospitals to collaboratively train ML models without sharing sensitive patient data. - Implemented XGBoost-based disease prediction models for conditions like diabetes, stroke, anemia, and breast cancer, improving predictive insights across distributed datasets. - Developed a full-stack web application using Flask, SQLite, and Gemini AI integration for health predictions, chatbot-based health tips, and appointment management. - Designed federated model aggregation using FedAvg algorithm to combine hospital-specific model weights securely while maintaining data privacy.

S

Skillyn - Smart ATS Resume Analyzer

Aug 2025

- Technologies: Python, Streamlit, NLP, Machine Learning - Built a platform to analyze resume-job description compatibility using AI. - Provided personalized suggestions to improve ATS-friendliness of resumes. - Designed an interactive and user-friendly interface with Streamlit. - Leveraged NLP techniques to extract key skills and match them with job requirements. - Helped users make the job search data-driven and more confident. - Deployed as a live web application: https://skillyn.streamlit.app/

R

Remedy Relay

Aug 2025

Remedy Relay is an AI-powered medical assistant designed to answer health-related queries using a structured knowledge base. Built with Flask, it leverages the Gemini API for natural language understanding and Pinecone for vector-based knowledge storage, enabling fast and relevant information retrieval. The current knowledge base is sourced from “The GALE ENCYCLOPEDIA of MEDICINE, Second Edition”, and while it’s small, the project is designed to scale with larger datasets over time. Remedy Relay demonstrates how technology can support education, quick reference, and accessible medical information. Key Highlights: - Built with Flask for backend processing. - Utilizes Gemini API for intelligent query responses. - Implements Pinecone vector da

A

Amazon Prime Videos Analysis

Jul 2024 - Aug 2024 (2 months)

- Conducted an in-depth analysis of Amazon Prime Video’s content library, visualizing data from 9,655 titles, spanning from 1920 to 2021, to uncover trends and patterns using Power BI. - Identified and categorized 519 unique genres, with Drama leading at 986 titles, enhancing content strategy insights. - Analyzed rating distribution revealing 2,100 titles in the 13+ category, aiding in understanding age-appropriate content distribution.

S

Stock Prediction Model

Jun 2024

- Developed an advanced web application using Streamlit to predict stock prices, integrating three cutting-edge models: LSTM, GRU, and TCN, resulting in a 15% improvement in prediction accuracy. - Enhanced model predictions by incorporating Exponentially Weighted Moving Average (EWMA) and Holt smoothing methods with spans of 50, 100, and 150 days, reducing Mean Squared Error (MSE) by 20%. - Utilized Yahoo Finance data through the yfinance module in Python, ensuring accurate and up-to-date financial data, leading to more reliable prediction outcomes.

P

Power BI Enabled Comprehensive Analysis on Unicorn Businesses

Feb 2024 - Mar 2024 (2 months)

Utilized Microsoft Power BI and advanced visual tools to analyze a comprehensive dataset of 1072 unicorn companies worldwide. Uncovered key insights into investment patterns, industry trends, and growth dynamics within the unicorn ecosystem. Notable findings include an average 7-year timeline to unicorn status, with a total funding pool of $592 billion and a valuation exceeding $4000 billion. Excited to contribute to the understanding of high-growth startups and venture capital.

E

EdMe (School Admission Website) Prototype

Jan 2024 - Apr 2024 (4 months)

We've developed a prototype for EdMe, a revolutionary platform empowering parents to make informed decisions about their children's education. EdMe offers comprehensive school discovery, comparison tools, expert counseling, and ensures authenticity in school information.

V

Volume Controller

Jan 2024

This project utilizes cutting-edge technologies, including #MediaPipe for precise hand tracking, #OpenCV for real-time computer vision, and #PyAutoGUI for seamless automation. With just your webcam and a few hand movements, you can now control your computer's volume effortlessly.

M

Music Player

Jan 2024

This Python-based Music Player Application offers a simple yet functional interface for playing, pausing, and controlling music tracks. Developed using Python's tkinter and pygame libraries, it allows users to select and play .mp3 audio files, adjust volume levels, and manage playback seamlessly. Whether you're a beginner exploring GUI development or seeking a customizable music player, this application provides an easy-to-use platform for enjoying your favorite tunes.

A

Advance Password Manager

Jan 2024

The project is a Personal Password Manager developed using Python's tkinter library, offering a secure and user-friendly interface for managing passwords. The application allows users to add, remove, edit, and view passwords for various websites, enhancing convenience and security in password management. Through the integration of dialog boxes and message boxes, the program ensures interactive and intuitive user experience, guiding users through password-related tasks effectively. Additionally, the project utilizes file handling to store passwords locally, prioritizing data privacy and accessibility. With features such as password encryption and a security key for accessing saved passwords, the Personal Password Manager reflects a commitmen

T

Turtle Crossing Game

Nov 2023

B

Bus Management System

Aug 2023 - Nov 2023 (4 months)

We designed a prototype (using Figma) and basic GUI (using python Tkinter module) to make travel easy of passengers by displaying current location of desired bus they want to get in and the approximate time the bus will take to reach the passenger's location. The prototype and GUI also have a workflow representing for admin and bus conductor. Admin have the options to add bus, view bus, manage conductors, edit bus details and remove bus. Conductor have the option to update the seat details and check if bus is full or seat are available. This project was completed under guidance of my Software Engineering Methodology's sir.

D

Documentation AI

A web application that automatically generates project documentation from a ZIP file containing source code. The platform analyzes the codebase and produces documentation in both Markdown and PDF formats, making it useful for students, hackathons and fast moving teams. Deployed on Vercel with a smooth and simple user experience. Key Features: • Upload ZIP and auto generate documentation • Download docs in MD and PDF formats • Fast processing and shareable output • Helps teams maintain clarity during collaboration

Vidhi Bhutia

About

Detail

Timeline

Résumé