AshleyLopez

• Developing a semantic segmentation pipeline in Python to extract solar panel boundaries from airborne RGB imagery captured at variable altitude, across uneven terrain and changing seasons.
• Assembling and validating a training set from 400+ partially annotated aerial images, and prototyping model-assisted annotation to generate examples for defect classes too rare to show up in production data.
• Presenting findings, data quality limitations, and technical tradeoffs directly to the client's Director of R&D.

• Designed and automated two end-to-end data pipelines across JMP, Statistica, and Spotfire, collapsing a 15-step manual workflow into a single repeatable run for the manufacturing analytics team.
• Structured and queried manufacturing datasets in Snowflake, letting five analysts pull batch data without engineering support.
• Worked with manufacturing scientists to turn analytical requirements into production workflows, documenting each one so it stayed in use after the co-op ended.

• Managed and audited facial recognition input data supporting statewide law enforcement training.
• Served as Google Workspace administrator, standardizing account provisioning across 20 classrooms and improving data accessibility for training programs.
• Built recurring reporting processes that cut manual requests and shortened setup time for new training cohorts.

• Diagnosed hardware and software failures across every College of Media Arts classroom using system logs and technical diagnostics, coordinating upgrades with engineering staff to improve reliability.

• Organize project data, estimates, and invoicing for the family business using digital tracking and communication tools.
• Develop online customer acquisition through Google Business and social campaigns.

• Architected a fully normalized (3NF) relational database modeling books, editions, authors, publishers, and sales transactions.
• Wrote multi-table joins and aggregations to evaluate sales performance and pricing strategy across several dimensions.

• Built a model to predict box office revenue from real-world film datasets, comparing feature sets and evaluating where the model's estimates broke down.

• Preprocessed 60,000+ applicant records with TF-IDF vectorization and feature engineering to retrain a classification model on a trusted data subset.
• Evaluated model behavior through correlation analysis and metric visualizations, surfacing the features driving unreliable predictions.

• Designed and implemented GUI gameplay on an MVC architecture for modular, testable logic.
• Collaborated on feature development using object-oriented and list-based structures.

Minor in Management Information Systems. GPA 3.62. Expected May 2027.

Coursework in machine learning, information retrieval, algorithms and data, and data management in business.

Honors: Huntington 100 · Dean's List · Ujima Scholarship · Hispanic Scholarship Fund.