Ben Howard Email & Phone Number

Washington, Dc, Us

Pasadena, California, Us

Led development for a warehouse optimization application which was awarded a SBIR phase III.

San Jose, California, Us

- Built data models and consumption tooling using BigQuery, dbt, and superset- Assisted stakeholders across the business with accessing and interpreting company data- Responsible for ad-hoc data science tasks (e.g. prototypes, estimation models, pricing exercises)

Mclean, Va, Us

- Principle investigator for a research initiative on anomaly detection for tax network data using deep learning graph neural network (GNN) algorithms.- Technical lead for IRS contract applied genetic algorithms to tax law in order to detect emergent issues.

Mclean, Va, Us

- Technical lead and deputy project lead for large IRS analytics contract. Helped shaped project goals, organized weekly meetings with the government task manager, wrote work plans and deliverables, and coordinated efforts through biweekly meetings for the project’s five distinct subtasks.- Authored ETL, modeling, and validation pipelines for supervised machine learning algorithms and inferential modeling. - Primary developer of API web service for training models. Functionality included cross-validation routines, state-saving, prediction using trained models on new data sets, and deployment on a NiFi model platform. Algorithms include XGBoost, Support Vector Machines, Random Forests with the functionality to extend to new algorithms. API was developed using Flask and service was hosted on an OpenShift cluster using Docker containers.- Developed query tool that enabled different tasks to collaborate on SQL code

Mclean, Va, Us

- Developed supervised machine learning models and data ETL pipelines for the IRS.- Made recommendations on how to improve the quality of an estimator from a stratified sample for a major IRS program.- Developed graph anomaly detection algorithms in physician transaction records for CMS. Served results in Tableau.

Mclean, Va, Us

Developed a prototype visualization for Latent Dirichlet Distribution (LDA) topic modeling using D3js.

Mclean, Va, Us

Designed and implemented an ontology RDF store for a web app.

Implemented a data profiler, and MySQL Load Data inFile and GATE NLP components for the Twister ETL tool.

Rochester, Ny, Us

Master Of Science (Ms), Statistics

Bs, Mathematics