Projects
A collection of projects spanning machine learning, data engineering, and full-stack development. Each project showcases production-ready solutions to real-world problems.
Energy Analytics for Sustainable Manufacturing
Developed analytical workflows for WVU’s Pollution Prevention Lab to evaluate energy usage across HVAC, lighting, motors, and process equipment. Built models to estimate kWh and fuel savings, greenhouse-gas reductions, and ROI/payback for REAP grants. Conducted on-site audits, created structured datasets, and delivered data-backed recommendations that improved energy efficiency for small and medium-sized businesses.
Security Log-to-Control Mapping System
As part of Data Driven WV’s experiential learning program (under NDA), contributed to a secure system that ingests, parses, and classifies logs to map events to relevant security controls. Designed components for structured log processing, anomaly identification, and compliance-aligned categorization. Focused on reliability, data quality, and building logic layers that support faster detection and actionable reporting while following strict security guidelines.
Autonomous Mobile Robot Path Planner
Developed a comprehensive Python framework enabling autonomous mobile robot (AMR) navigation on grid maps. Supports classic search algorithms (A*, Dijkstra), dynamic obstacle handling, path smoothing (Bezier, spline, shortcut), and real-time simulation. Modular architecture allows plug-and-play of planning algorithms and movement models for rapid prototyping.
La Liga Player Performance Analysis (PCA, 2023/24)
Applied Principal Component Analysis (PCA) on ~3,600 match-records (≈681 players) from La Liga 2023/24. Reduced 150+ features into principal components to reveal hidden structure in playing styles—identifying archetypes like ball-playing defenders, creators, finishers, and high-consistency players. Demonstrates dimensionality reduction for sports analytics and talent evaluation.
Heating Tunnel Operational Efficiency ML Analysis
Performed feature engineering, statistical diagnostics, and ML modeling on industrial heating-tunnel data. Identified inefficiencies and recommended optimal process settings to reduce energy use while maintaining throughput. Demonstrates real-world industrial analytics and manufacturing optimization.
Automated Defect Detection System for Manufacturing QA
Developed a computer-vision based system to detect surface flaws, misalignments, and anomalies in manufacturing outputs. Built a modular ML pipeline for training, classification, and evaluation using high-resolution images. Reduces manual inspection load and improves defect detection accuracy for high-throughput production.
Health Data Analytics Pipeline
Built a reproducible analytics pipeline with Airflow orchestration, dbt modeling, PostgreSQL warehouse storage, Dockerized deployments, and Superset dashboards. Cleans, validates, transforms, and visualizes breast-cancer dataset records to extract clinical insights.