William Patric Poston

Old Dominion University Research Foundation — Norfolk, VA

• Develop scalable geospatial machine-learning pipelines using Google AlphaEarth Foundations' 64-dimensional satellite embeddings to analyze land and environmental patterns across Hampton Roads, Virginia.
• Process 64-band, 10-meter-resolution raster imagery through Google Earth Engine and ODU's Waterfield high-performance computing cluster.
• Constructed an analysis-ready dataset containing 99,986 geospatial observations and 71 variables, including spatial metadata and 64 learned embedding dimensions.
• Apply exploratory analysis, PCA, k-means clustering, and geospatial visualization to interpret structure in high-dimensional satellite data.

KBR

• Supported test and evaluation of satellite communications systems for the U.S. Air Force and Space Force Nuclear Command, Control, and Communications (NC3) program.
• Analyzed system behavior and test results to identify failure modes and evaluate communications reliability under simulated operational conditions.
• Collaborated with multidisciplinary engineering and analysis teams in a mission-critical, security-controlled environment.

NASA Artemis II Project

• Led a 10-member multidisciplinary engineering team developing an experimental lunar satellite sensor system for the Artemis II program.
• Coordinated hardware-software integration and designed, programmed, and tested sensors measuring seismic activity, temperature, gravity, and magnetic fields.
• Validated end-to-end collection and transmission of spacecraft sensor data to a simulated Mission Control environment.

CoVA CCI

• Developed a subscription-based browser extension that protects users from malware and other cyber threats while promoting safe browsing practices.
• Worked autonomously within a peer development team to translate stakeholder needs into a functional web product.
• Won First Place in the 2022 INNOVATE Cyber Competition and secured continued stakeholder sponsorship for system operations.

• Developed a constraint-based backtracking engine that solves and generates crossword puzzles using ENWiktionary, a dictionary containing more than 10 million entries.
• Implemented candidate filtering and recursive search logic to satisfy intersecting-letter constraints and reduce the solution space.

• Developed a personalized travel-planning application that generates day-by-day itineraries using scraped destination data and generative AI.
• Designed a hybrid recommendation engine that ranks activities according to user preferences before supplying structured candidates to an OpenAI-powered generation workflow.

• Implemented a feedforward neural network and training pipeline from scratch in Python without relying on a deep-learning framework.
• Created a Tkinter interface that lets users draw digits, perform inference, train custom models, and load saved model parameters.

Master of Science (MS), Computer Science — GPA: 4.00/4.00

Master of Science (MS), Data Science — Concentration in AI & Machine Learning — GPA: 4.00/4.00

Bachelor of Science (BS), Computer Science — Minor in Data Science — Perry Honors College Graduate — GPA: 3.55/4.00

Associate of Science (AS), Computer Science — GPA: 3.89/4.00