Glen Smith

Virginia Tech Applied Research Corp.

Aug 2020 - Present

Lead a research project to develop a novel search framework in Python by using NLP techniques to embed unstructured text documents into multi-dimensional numerical vectors and project those vectors into a 3D space, allowing for exploration and analysis.
Developed a backend API using Flask and ElasticSearch in Docker to validate requests, build and execute queries on the database, process results, and generate analysis data for front end web service consumption.
Designing the corporation’s Big Data Architecture solution by employing enterprise-level data storage and search frameworks such as Hadoop, Spark, and ElasticSearch to centralize data collection efforts across our 3 divisions.

Virginia Tech Applied Research Corp.

Jan 2020 - Aug 2020

Developed a company search application using R-Shiny (R) and ElasticSearch to cut our technology scouting market analysis timeframe in half and provide analysis and exploration opportunities via geographical and funding trend visualizations.
Utilized statistical methods such a t-test, ML models (logistic regression, random forest), and feature engineering on 8 years of USAF pilot training data to predict pass/failure and presented findings to AFB officials and instructors to allow them to identify shortcomings early and reduce spending costs on overtraining.
Built a web scraping tool in Python to walk company websites and collect additional “about me” information to populate our CRM database which reduced our description deficit from 55% to 30%.

BCPS Office of Media and Instructional Tech.

2016

Enabled real-time data analysis across the organization by developing an online database for hosting public school & other non-sensitive data utilizing Google Cloud SQL and Python.
Wrangled and compiled data from multiple sources, including school webpages, Excel, and text files.
Developed a website for the online database and assisted in the UI design using HTML.

Research Summary