Projects
A collection of personal and professional projects showcasing backend development, machine learning, and full-stack engineering.
Filter by:
AllFull-StackML/AIRoboticsLLMData
Showing 8 of 8 projects
AvaDocs — AI Research Assistant
FeaturedRAG-powered document analysis with a personality
Problem: Generic chatbots hallucinate and can't be trusted to answer questions about specific documents.
Solution: Built a complete RAG pipeline on Django: PDFs are ingested, chunked, and embedded into a vector store.
Impact: Deployed publicly as an interactive portfolio demo.
Next.jsTypeScriptReactDjangoPostgreSQLRedis+3 more
2025•LLM•Personal Project
CERN Particle Physics ML Pipeline
FeaturedNeural network pipeline for ATLAS detector signal separation
Problem: CERN's ATLAS detector generates massive amounts of collision data, but signal events (interesting physics) are buried in background noise.
Solution: Developed a complete neural network pipeline in Python using Keras and scikit-learn.
Impact: Successfully trained neural network to separate signal from background in ATLAS collision data.
PythonKerasTensorFlow
2024-2025•ML/AI•California State University, Stanislaus
NASA Lunar Robotics Challenge
FeaturedAutonomous lunar surface navigation with computer vision
Problem: NASA's lunar challenge required autonomous robots to navigate the lunar surface, detect and classify rocks, and make intelligent decisions without human intervention.
Solution: Set up custom Carla simulation environment and managed team development workflow.
Impact: Successfully submitted working solution to NASA competition.
PythonDockerCarla
2025•Robotics•NASA Lunar Challenge
Probability-Based Map Fusion — Research Paper
Published method for fusing occupancy maps with neural network object labels
Problem: SLAM-based robotic mapping systems produce geometric representations of the environment but lack semantic understanding of what objects are present and where.
Solution: Developed a probability-based fusion algorithm that aggregates neural network object detections over time, weighting observations by distance from stored map points using radial basis functions and the midpoint circle algorithm for efficient spatial cell filtering.
Impact: Demonstrated that lightweight, probability-based methods can produce richer, context-aware maps that support object search, counting, and location estimation — improving robot decision-making without sacrificing performance.
PythonROS2Keras
2025•Robotics•California State University, Stanislaus
ROS2 Neural Network Robotics
Sensor fusion and spatial reasoning with neural networks
Problem: Robots need to understand not just what objects are present, but where they are in 3D space.
Solution: Built ROS2 environment for sensor data fusion.
Impact: Demonstrated advanced robotics concepts combining perception (CNN), localization, and spatial reasoning.
PythonROS2
2025•Robotics•California State University, Stanislaus
eTrademark Processing Platform
FeaturedFull-stack trademark application processing on AWS
Problem: Trademark attorneys needed a streamlined way to process trademark applications, but existing solutions were slow, expensive, and didn't integrate well with USPTO data sources.
Solution: Designed and built a full-stack web application with React frontend and dual-database backend (MongoDB for document storage, PostgreSQL for relational data).
Impact: Successfully deployed to production serving real trademark processing clients.
ReactJavaScriptMongoDBPostgreSQLAWSEC2+2 more
2020-2023•Full-Stack•eTrademark
Animated Hero Banner — itslilym.com
Custom JavaScript animation for a personal creative website
Problem: The client wanted her personal website to open with an animated hero section that felt polished and unique — something that standard website builder templates couldn't provide out of the box.
Solution: Built a custom JavaScript animation using tweening techniques to orchestrate smooth transitions, easing curves, and motion effects in the hero section, integrated directly into her existing custom website setup.
Impact: The client was delighted with the result.
JavaScript
2021•Frontend•itslilym.com
MCSquared Health Platform
HIPAA-compliant health management application
Problem: Health startup needed a web application to manage patient data and health tracking, but had to ensure strict HIPAA compliance for data security and privacy.
Solution: Built full-stack React application hosted on Heroku.
Impact: Application helped startup raise over $100,000 in seed funding.
ReactJavaScriptNode.js
2019-2020•Full-Stack•MCSquared Health