Projects

illustrations illustrations illustrations illustrations illustrations illustrations illustrations
post-thumb

GraphQL @ Chevron

Used Hasura to give organizational users API access to MarkLogic databases as well as 26 other data stores in Azure. Petroleum operators now have 360-degree views of equipment repair histories in a variety of tools including Power BI and Tableau.

Project completed on 4/3/2020


post-thumb

Database Performance @ Lockheed Martin

Provided database performance modeling and data architecture assessment for F-35 Joint Strike Fighter program. Compared leading database technologies including PostgreSQL, SQL Server, MongoDB and Neo4j to recommend a suitable vendor for their mission planning platform.

Project completed on 9/15/2020


post-thumb

Cash Register Metrics @ Macy's

Having operational cash registers are paramount for a large retailer like Macy's. Leveraging open source technologies such as Telegraf, InfluxDB and Grafana, I created a time series database that will alert Macy's IT staff to cash register issues in under 5 seconds.

Project completed on 11/1/2020


post-thumb

Analytics @ AT&T

Supported multiple projects at AT&T to provide executive decision makers with real time reporting on sales, promotions and customer churn. This was achieved by developing a large Oracle data warehouse, developing predictive models in R & SPSS and creating dashboards in Cognos.

Project completed on 8/15/2014


post-thumb

IoT @ Xylem

Xylem (NYSE: XYL) is a large manufacture of water meters and valves. I worked with their team of developers to migrate the backend of a smart water meter historian system from MongoDB to InfluxDB which resulted in a cost savings that was large enough to hire 4 additional staff.





post-thumb

Flight Data @ Naval Air Systems Command

The Naval Air Systems Command provides full life-cycle support of naval aviation aircraft. This includes storage and analysis of flight data from Joint Strike Fighter (JSF) which generates over 1TB of data per hour of flight. We developed a pipeline to move this data from the aircraft at the edge and into a Hadoop data lake. Additionally. we created a separated ETL tool using Presto to load aggregated flight into a Teradata data warehouse to enable data scientists to expand the useful service life of various aircraft parts.



post-thumb

Machine Learning @ Great West

Great West Casualty is a large insurer in the commercial trucking industry. They had an important predictive model in R that as taking 7 days to compute because it looped 300M records. By translating the R code to Python, preprocessing the records in a new data mart and refactoring the code, I got the runtime down to just 45 minutes which allowed actuaries to explore more scenarios and assumptions.