Machine Learning & Data Science Intern

Madeline Blount

503-***-**** - *************@*****.*** - La Jolla, CA

Data science and machine learning practitioner with experience building scalable data pipelines, developing predictive models, and analyzing large, unstructured datasets. I work with cross-functional teams to integrate AI and machine learning into work processes. Education

UC San Diego, September 2022 - June 2026

Bachelor of Science, Cognitive Science with a Specialization in Machine Learning. Current GPA: 3.74 Provost Honors: Fall 2022, Winter 2023, Spring 2023, Fall 2023, Winter 2024, Fall 2024 Coursework: Artificial Intelligence Algorithms, Neural Networks and Deep Learning, Introduction to Machine Learning, Linear Algebra, Data Science in Practice, Probability, Vector Calculus, Statistics Technical Skills

● Programming languages: Python, Java, C, Matlab, R, SQL, Jamovi

● Data science and ML/AI library frameworks: NumPy, Pandas, Matplotlib, Seaborn, Scikit-learn, PyTorch, NLTK, NLP

● Communication and software: Git and GitHub, Microsoft, Google Suite, Slack Laboratory and Internship Experience

ML Programmer and App Developer, Delta Rising Foundation, Aug 2025 - Present

● Engineered end-to-end pipelines for cleaning and analyzing satellite and climate datasets across 6 counties.

● Developed LightGBM predictive model for carbon sequestration using regression and ensemble methods, achieving an r-squared value of 0.91.

● Improved resolution of the model from 64 square meters to a higher resolution of 25 square meters. Program Developer of AI Agent Video Analyzer, Serious Games at Qualcomm Institute at UCSD, Sep 2025-May 2026

● Built scalable data pipelines to process and integrate multimodal video data for real-time analysis.

● Incorporate AI to analyze Unreal Engine video streams using YOLO computer vision models.

● Partnered with cross-functional teams to define areas for improvements and deliver on them. Market Researcher and Ad Developer, Delta Rising Foundation, May 2026 - Present

● Conduct market research into ICPs to inform app development decisions and ad development.

● Analyze ad performance and iteratively adjust ads to increase engagement. Programmer and Lab Intern in Nitz Systems Neuroscience Laboratory of UCSD, Aug 2024 - Present

● Developed automated labeling program (MATLAB) achieving over 99% accuracy on behavioral datasets.

● Integrated data analysis program to reduce time spent labeling to a tenth of previous time spent.

● Cleaned, structured, and managed large relational datasets for ML applications.

● Performed EDA and statistical modeling on high-dimensional neural data. Data Analyst and Financial Advisor, Triton Television (UCSD), April 2024 - Present

● Analyzed financial data for KPIs and created data visualizations and dashboards.

● Delivered data-driven presentations to non-technical stakeholders, increasing allocated budget from $35K to $55K. Machine Learning and Data Science Projects

Building Predictive Models via NLP & ML on Sephora Dataset, Sept 2024 - Dec 2024

● Collected, cleaned, and merged large-scale product review datasets using Pandas and NumPy.

● Applied NLP techniques (tokenization, sentiment analysis with NLTK) to extract structured features.

● Built predictive models to recommend products based on user attributes and preferences with over 90% prediction accuracy.

● Performed exploratory data analysis and correlation analysis to identify key drivers of product performance. Emotion Classification from Audio (PyTorch) Feb 2025 - March 2025

● Developed CNN-based models achieving 75% accuracy on emotion detection.

● Built data pipelines using Google AudioSet and visualized model performance. Machine Learning Programmer, Image Classification and Reconstruction of Simpsons Characters, Sep 2024 - Dec 2024

● Pre-processed images to prepare for machine learning applications and data analysis.

● Applied the unsupervised machine learning technique, principal component analysis (PCA), to extract key features, then used clustering and classification methods to identify character images, attaining at least 80% accuracy.

Contact this candidate