About Me
I'm a Data Scientist and Machine Learning Engineer with a passion for building intelligent systems that solve real-world problems. My expertise lies in developing and deploying machine learning models, with a particular focus on computer vision and natural language processing.
I'm currently working on projects that combine my interests in sports analytics and machine learning, developing tools that help teams make data-driven decisions and gain competitive advantages.
Education
University of Southern California
Master of Science in Applied Data Science
August 2024 - May 2026
Viterbi School of Engineering
GPA: 3.86
Chapman University
Bachelor of Science in Data Analytics
August 2020 - May 2024
Fowler School of Engineering
GPA: 3.86
Minor in Business Administration
Professional Experience
Machine Learning Engineer
PushButton.AI - Course Building Program • Orange, California
January 2024 - Present
- Developed and deployed LLM APIs with Docker and Redis on DigitalOcean, reducing API costs by 60% while ensuring scalability for high-load applications.
- Evaluated and optimized open-source HuggingFace LLMs on CUDA cloud hardware, balancing cost, latency, and accuracy for real-time AI solutions.
- Implemented high-token-output RAG systems using recursive calls and prompt engineering, increasing model output capacity by 5× and retrieval efficiency.
- Integrated web scraping with RAG workflows, enhancing document retrieval quality and increasing customer retention by 15%.
- Designed and refined LLM prompts to generate structured 10,000-word program outputs as functional HTML, ensuring readability for 6th to 12th grade levels.
- Automated agentic workflows to convert course outlines into 12 structured lessons, integrating task IDs and real-time tracking for dynamic content generation.
- Stress-tested API infrastructure to handle peak loads, optimizing request handling and minimizing downtime for high-traffic applications.
Co-Founder | Cloud Engineer
Ovrsee
January 2024 - October 2024
- Develop API services via Google Cloud Functions handling TikTok OAuth workflows, processing 48+ daily requests per user
- Streamline user metric aggregation (daily/weekly/monthly) using Firebase with automated Cloud Task scheduling - achieving 90% front-end speed improvements
- Integrate API operations with Firebase to manage user authentication while structuring database storage systems for 3 customers: brand owner, manager, and creator
- Enforce encrypted key management and validations for API requests, safeguarding TikTok API access and user data integrity
Data Scientist Internship
Cabrillo Marine Aquarium • San Pedro, California
June 2023 - August 2023
- Gained unique experience working with TRACKS data software, inputting, and altering animal data recorded from aquarist activity
- Visualized enrichment and welfare animal statistics for the aquarium to display around exhibits
- Set up semi-auto generation for visualizations for aquarists to access post internship period
- Learned basic feeding, cleaning, and upkeep techniques to assist aquarists with animal care around the facility as needed
My Projects
PLG SaaS Marketing Mix Modeling
• Built a full Bayesian MMM pipeline in R with adstock decay, Hill saturation, and hierarchical priors to analyze PLG SaaS marketing impact • Simulated realistic multi-channel data with seasonal trends and business logic to validate model assumptions and inference quality • Used brms and Stan to run MCMC sampling with custom priors and diagnostics, ensuring robust, reproducible statistical modeling • Automated visual reporting for ROI, channel attribution, and diagnostics, translating complex outputs into clear marketing insights
ChatDB - NBA SQL RAG
• Created a natural language–to–SQL interface using OpenAI, mapping conversational questions to structured queries with a focus on accessibility and transparency • Built a Streamlit and MySQL interface with query tracking and live data visuals, reducing barriers for non-technical users and demonstrating empathetic product thinking • Implemented secure database modification with rollback and validation, emphasizing trustworthiness and long-term maintainability • Designed a modular, well-documented system with onboarding in mind, reducing dev ramp-up time and showing commitment to team success
TikTok Virality Predictor
Trained a computer vision model built on ResNet (2+1)D leveraging PyTorch to predict virality from TikTok video input. Wrote web scraping scripts to download and process TikTok videos using Python and Pandas to engineer virality score.
DrugAI-CVAE
Built and trained an Autoregressive CVAE in PyTorch to generate novel drug molecule sequences (SMILE) based on target proteins.
Human Activity in VR Research
• Designed 13 modular functions to generate advanced 3D models and visualize VR hand/object trajectories through dynamic line charts • Leveraged movement pattern visualizations to design novel rule-based system for identifying and labelling user actions • Implemented unsupervised ML models analyzing 8 movement trajectory types for VR activity prediction using spatial pattern
NBAR
This project provides a robust framework for analyzing NBA player performance data against betting odds. It features automated data collection from official NBA box scores, integration with betting odds APIs, and statistical modeling to identify patterns and potential value opportunities in the sports betting market.
Pest Patrol
Implementation of detect-net software on NVIDIA Jetson Nano to detect and spray pests like Raccoons for protection.
Fire Segmentation
U-Net model trained on a custom dataset of fire and non-fire images to segment fire in real-time video streams.
Psyche Evaluation
Front-end and back-end development of a SQL application to evaluate psyches and provide analyses on scores. The application includes Big-5 and Narcissism tests.
Money Wicket
MoneyWicket is a representation of modern sports data analysis applied to the game of Cricket. It utilizes Logistic Regression and regularization models like Ridge and Lasso in an attempt to create predictive models that could be used to help scout possible transfers and minimize money expenditures (only buying good players for the team).
Soccer Linear & Logistic Analysis
Personal class project where I collected a dataset of 2018-19 English Premier League statistics and used python + pandas to put the data through different tools and models.
Skills & Technologies
Programming Languages
- Python
- R
- SQL
- HTML
Data Analysis & Visualization
- Pandas
- ggplot
- Plotly
- NumPy
- Matplotlib
Machine Learning
- Keras
- Tensorflow
- PyTorch
- LLMs
- NLP
- Langchain
- Computer Vision
- Deep Learning
Software Deployment
- Google Cloud Projects
- Docker
- git
- Amazon Web Services
- Firebase
- REST API
- Flask
- Redis
- Cloud
Honors & Awards
Provosts List
Chapman University
GPA above 3.8 for a semester (2020-2024)
Program Honors
Chapman University
Excellence in Computing Studies
Association for Computing Machinery