Transforming Data into Intelligent Decisions

A Portfolio Showcasing My Academic & Professional Journey

Harsh's Bio

I’m a graduate student at NYU SPS pursuing an MS in Management & Analytics, with a background in Computer Engineering. I work at the intersection of AI/ML, data analytics, and enterprise risk management, with hands-on experience building real-world systems across machine learning, computer vision, and data-driven applications

View Portfolio

Education

B.Tech Integrated Computer Engineering

I completed my undergraduate studies in Computer Engineering at Mukesh Patel School of Technology Management & Engineering, building a strong foundation in programming, data structures, databases, and machine learning through hands-on, data-driven projects and applied coursework.

MS in Management & Analytics

I am currently pursuing an MS in Management & Analytics at NYU School of Professional Studies, where my studies focus on data analytics, machine learning applications, and enterprise risk management. Through coursework and applied projects, I work on translating complex data into actionable insights, bridging technical analytics with business strategy and risk-aware decision systems.

Research & Publications

Research & Publications

My Projects

Welcome to my portfolio. Here you’ll find a selection of my work. Explore my projects to learn more about what I do.

Multi-Agent Vendor Risk & Due-Diligence Platform

Five specialised AI agents independently investigate a vendor — financial, compliance, operational, reputation, and strategic fit — then a meta-agent fuses their findings into a single auditable Approve/Review/Reject recommendation with evidence, confidence scores, and red flags. Built on IBM WatsonX with LangChain.

Risk Radar — AI Project Risk & Decision Engine

An AI-powered project intelligence tool that turns a raw brief and team roster into a skill-matched plan, a live 0–100 risk score, and a ranked list of the decisions a manager must make before a project breaks.

Customer Segmentation & Behavioural Analysis (SPSS)

Analysed large-scale behavioural datasets using SPSS to identify patterns, trends, and user segments
Performed statistical analysis (correlation, segmentation techniques) to understand key drivers of user behaviour
Applied data cleaning and validation techniques, improving data accuracy and reliability by 15%
Interpreted results and translated findings into clear, structured insights for non-technical stakeholders
Created visualisations to communicate behavioural trends and support data-driven decision-making

Enterprise Automation with Microsoft Power Automate

Designed and implemented business process automation solutions using Microsoft Power Automate and Power Apps to replace manual, spreadsheet-based workflows. Built automated forms with validation logic, integrated SharePoint lists for structured data management, and developed cloud flows for approvals, notifications, and event-based triggers. The solution improved data accuracy, reduced manual effort, and enabled non-technical users to manage workflows through intuitive interfaces, demonstrating the impact of low-code automation on operational efficiency and scalable enterprise processes.

College Compass

College Compass is a data-driven university recommendation platform that helps students identify best-fit colleges based on academic profile, preferences, and interests. The system integrates machine learning and NLP techniques to analyze structured and unstructured data, generate personalized recommendations, and improve decision-making through continuous evaluation and refinement.

Pneumonia Detection with Chest X-rays

Pneumonia Detection with Chest X-rays is a medical imaging application that uses deep learning models to automatically detect pneumonia from chest X-ray images. The system applies convolutional neural networks (CNNs) to analyze radiographic patterns, support faster diagnosis, and improve accuracy through model evaluation and optimization.

Stock Price Prediction

Reinforcement Learning–Based Stock Trading applies RL algorithms such as PPO, DQN, and SAC to learn optimal buy, sell, and hold strategies from historical stock market data. Using Yahoo Finance data and Stable-Baselines3, the system adapts to changing market conditions and identifies profitable trading patterns with minimal human intervention.

Car Price Prediction

Car Price Prediction is a machine learning application that predicts the market price of used cars based on features such as brand, model, year, mileage, fuel type, and engine specifications. The system applies regression techniques to analyse historical data, identify key pricing factors, and generate accurate, data-driven price estimates.

Cryptocurrency Dashboard

Cryptocurrency Dashboard is an interactive data analytics application that visualizes real-time and historical cryptocurrency market data. The dashboard presents price trends, volume, and performance metrics through intuitive charts, enabling users to analyze market movements and make data-driven insights into digital asset behavior.

Smile Foundation- NGO website for children

NGO Website for Children is a full-stack web platform designed to connect donors with NGOs supporting underprivileged children. The application enables secure donations, real-time NGO discovery using location services, and transparent tracking of contributions to improve accessibility, trust, and impact in charitable giving.

.pro-gallery-wix-wrapper {display: block !important;} .pro-gallery-wix-wrapper .gallery-item-container {opacity: 1 !important; display: block !important;}

Stay Updated with My Work and Insights

Get in Touch

Your feedback and inquiries are welcome!