Personal Projects

Self-Driving Car Simulation via Neuroevolution

AI & Machine Learning • Neuroevolution

Built a 2D autonomous driving simulation where cars learn to navigate procedurally generated tracks using Neuroevolution. Instead of traditional programming, neural networks evolve through a Genetic Algorithm, developing sophisticated driving behaviors such as smooth cornering and obstacle avoidance without any human-coded rules.

  • Neural Networks & Genetic Algorithm: Designed feedforward neural networks from scratch and implemented evolutionary training with fitness evaluation, selection, and mutation.
  • Sensor Simulation: Created a custom LiDAR-like sensor system using line intersection mathematics for real-time track boundary detection.
  • Physics & Optimization: Implemented 2D vehicle kinematics, vector calculations, and performance optimization to support 100+ concurrent AI agents.
  • Model Persistence: Enabled saving and loading of the best-performing "brain" using pickle for continuous learning across sessions.

Python, Pygame, Neural Networks, Genetic Algorithms, OOP, Physics Simulation

GeoHouse — AI-Driven Property Insights

Cloud-Native Geospatial Engineering

Built an automated due diligence platform that empowers users to evaluate property addresses instantly. The service aggregates multi-dimensional data and leverages AI to provide qualitative assessments on safety, market value, and neighborhood demographics.

  • Data Integration: Engineered a pipeline to fetch and normalize real-time data from property, crime, and census APIs.
  • AI Analysis: Implemented an LLM-powered evaluation engine to transform raw data into actionable investment and residency insights.
  • Scalable Infrastructure: Developed with FastAPI and Docker; deployed on Google Cloud Run using a CI/CD workflow for seamless scalability.

Python (FastAPI) • Docker • Google Cloud Run • Gemini/LLM • RESTful APIs

Professional Portfolio

Full-Stack Personal Project

Developed a high-performance personal portfolio to showcase engineering projects. This project highlights the integration of modern web technologies with AI-driven development workflows and professional cloud hosting.

  • AI-Enhanced Workflow: Utilized Generative AI for rapid UI/UX prototyping and code optimization, ensuring a clean, responsive user experience.
  • Backend Architecture: Implemented a Flask server to manage site structure, routing, and dynamic assets efficiently.
  • Live Deployment: Successfully managed production hosting, domain DNS records via Namecheap, and a Git-based workflow for continuous updates.

Flask • Python • Bootstrap • Git • GenAI

PA Municipality Vector Data Analysis

Geospatial Data Exploration Project

Developed a systematic geospatial analysis workflow to process and visualize Pennsylvania's municipal boundary data. This project serves as a practical implementation of vector data engineering, focusing on the accurate transformation of raw spatial geometries into meaningful geographic insights.

  • Geometric Attribute Calculation: Implemented precise CRS transformations to compute municipal areas in square kilometers, ensuring mathematical accuracy between geographic and projected systems.
  • Spatial Distribution Mapping: Created multi-layered visualizations, including static choropleth maps and interactive Folium web maps, to represent the administrative distribution and spatial density of regional boundaries.
  • Exploratory Spatial Data Analysis (ESDA): Utilized GeoPandas and Shapely to perform deep dives into polygon datasets, extracting key geometric features and handling spatial metadata.

Python • GeoPandas • Matplotlib • Folium • Shapely • EPSG Transformations

Work Experiences

FME Python Transformer for Utility Data

Tata Consultancy Services | Pacific Gas & Electric Company (PG&E)

Built custom Python transformers within FME to process complex water and electricity utility datasets for Pacific Gas & Electric Company. These transformers perform comprehensive data validation, transformation, cleansing, and integration tasks, ensuring high data quality and consistency across multiple systems. The entire workflow is fully automated and deployed on FME Flow, enabling scheduled, reliable, and hands-free execution. This solution significantly improved data processing efficiency, reduced manual effort, and enhanced the overall reliability of utility data integration into GIS and business applications.

FME • Python • FME Flow • ETL Pipelines • Automation

Enterprise ArcGIS Online Data Automation Pipeline

Tata Consultancy Services | Pacific Gas & Electric Company (PG&E)

Designed, developed, and maintained a scalable enterprise data pipeline for ArcGIS Online at Pacific Gas & Electric Company.

  • Designed and deployed automated data refresh pipelines using AWS Lambda and EventBridge
  • Implemented secure PostgreSQL connections using AWS Secrets Manager
  • Automated quarterly data synchronization between PostgreSQL and ArcGIS Online
  • Developed custom ArcPy geoprocessing tools for spatial data transformation and reporting
  • Created custom field forms and interactive pop-ups in ArcGIS Field Maps for field workers
  • Significantly reduced manual intervention and improved overall data accuracy and reliability

AWS Lambda • EventBridge • ArcGIS Online • ArcPy • PostgreSQL • FME • Agile

MEAN Stack Enterprise Training Modules

Full-Stack Development Project

Developed a comprehensive suite of full-stack applications utilizing the MEAN (MongoDB, Express.js, Angular, Node.js) ecosystem. This project demonstrates proficiency in building scalable web solutions, real-time communication systems, and automated deployment pipelines.

  • Full-Stack Integration: Engineered multiple modules including a Budget Planner, Cart Management system, and Blogging platform using Angular for dynamic front-ends and Node.js/Express for robust back-end logic.
  • Real-Time & Data Management: Implemented real-time messaging using Socket.io and established secure database connections with MongoDB for persistent log recording and task analysis.
  • DevOps & Containerization: Streamlined the development lifecycle by integrating Jenkins and Docker, ensuring consistent environment configuration and automated CI/CD workflows.

Angular • Node.js • MongoDB • Express • TypeScript • Socket.io • Docker • Jenkins


Project Demonstrations & Examples

Self-Driving Simulation in Action

Real-time neuroevolution training process showing agents learning to navigate.