Projects & Research

Predictive Modeling for Country Vulnerability Assessment

Hoover Institution, Stanford University, USA (Jul 2025 – May 2026)
Tech: Python, LSTM

• Design and implement predictive models using machine learning and deep learning for multi-horizon vulnerability estimation.
• Collaborate with interdisciplinary teams to integrate diverse country-specific data streams into robust modeling pipelines.
• Validate and refine models through historical analysis to improve accuracy, robustness, and actionable insights.

CNN-Based Classification of Tabular Data

Occidental College, USA (Aug 2024 – Jul 2025)
Tech: Python, CNN, Tabular-to-Image Transformation

• Transformed structured cancer datasets into image-like formats for CNN-based classification.
• Improved diagnostic accuracy for early-stage cancer detection by leveraging spatial encodings.
• Focused on structured semantics, model interpretability, and academic publication.

Pre-Natal Risk Assessment Engine

Tosoh India Pvt. Ltd. (Oct 2024 – Mar 2025)
Tech: Python, Neural Networks, XGBoost, Random Forest, Differential Privacy

• Built an explainable AI system to predict risks of congenital disorders (Down Syndrome, Trisomy 18/13, ONTD, SLOS) from maternal demographic inputs.
• Integrated ensemble models and differential privacy for clinical interpretability and data protection.
• Developed a robust API to enable middleware integration.

Stock Market Crash Prediction

Florida International University, USA (Apr 2024 – Jul 2024)
Tech: Python, Time Series, Supervised Learning

• Analyzed 10 years of stock price data to identify crash-like patterns using ML classification techniques.
• Engineered datasets using sequences of closing prices and developed predictive models using multiple supervised algorithms.
• Co-authored a research paper titled Forecasting Stock Market Crashes Using Deep Learning (under review).

Chromatography Data Analysis Platform

Entiovi Technologies Pvt. Ltd. (Dec 2023 – Jul 2024)
Tech: Prolog, Python, Cryptography

• Developed an intelligent interpretation engine for HbA1c chromatograms using symbolic logic in Prolog.
• Implemented encryption to ensure patient data privacy in diagnostic pipelines.
• Helped automate detection of hemoglobinopathies based on peak classifications and retention times.

VC Industry Research & Pitch Deck Analysis

Energy Innovation Capital, USA (Sep 2024 – Oct 2024)
Tech: Market Research, Strategic Analysis

• Analyzed trends in the renewable energy VC landscape.
• Evaluated real startup pitch decks and delivered investment summaries with competitive insights.
• Gained exposure to the intersection of venture capital, sustainability, and innovation strategy.