Adaptive & Robust Network Intrusion Detection

Developed a three-track intrusion detection framework combining context-aware Random Forest classification, reinforcement-learning-based adaptive thresholding, and robustness evaluation under noisy and incomplete network data.

Controlled Evaluation of Structural Inductive Bias in Multi-Omics Representation Learning

Investigated how biological network structure affects learned representations of high-dimensional breast cancer multi-omics data through controlled deep learning experiments and graph-based regularization.

Multi-Agent LLM System for Clinical Information Extraction

Developed an experimental multi-agent NLP pipeline for automatically extracting structured cancer registry information from free-text pathology and medical reports using open-source large language models.

NLP-Based Business Matching & Recommendation System

Developed NLP-based recommendation and matching systems to connect professional profiles using semantic similarity, contextual embeddings, and transformer-based representations.

Gene Expression & Pathway Analysis for Breast Cancer Treatment Response

Designed and implemented an end-to-end bioinformatics pipeline to identify pathways linked to treatment response in breast cancer patients by analyzing RNA-Seq data before and after treatment.

Deep Learning Method for COVID-19 Detection in Chest X-ray Images

As a final-year BSc project, developed and evaluated a deep learning model to classify COVID-19 positive vs. normal chest X-rays using a large-scale public dataset (~25,000 images).

Grouping Text Inputs Based on Textual Similarity

This Streamlit app takes in a csv input and analyzes the contents (tags) using cosine similarity score to output tag groups where the most similar tags are grouped together. The user can download the generated tag groupings and use them to match elements with particular tags to other similar tags.

Interactive Dashboard for Quarterly Event & Engagement Survey

During a short internship, I worked on data visualization to document the participation and engagement of employees during the company's quarterly event. The participation data and survey responses were collected through PowerApps and SharePoint, and the dashboard was created using PowerBI.

Interactive Dashboard for Company Categorization

A Tableau Dashboard categorizing and displaying companies (from an internal private database) based on: -Their parent industry and subindustries, -The certifications of their products and/or services, -The country they are based in.

Interactive Dashboard for Industry Classification

A Tableau Dashboard displaying a hierarchical structure for categorizing companies (from an internal private database) based on a parent industry and multiple subindustries. The dashboard includes filtered views based on parent/subindustry.

Personal Portfolio Using Streamlit

Using Streamlit, I put together an interactive personal website that serves as an "extended resume", documenting professional, academic and personal details about me.