Development of automated machine learning system used as the core driver of Automate Ads paid media campaigns. Design, development, testing, and scaling of the Backbone.js...more based front end web app as well as the Node.js back end API. This was accomplished using a multitude of different software languages, database solutions, big data technologies, and cloud hosting environments. The entire product was completed using a small team of 3 engineers.
Developed the processing systems of test data collected from multiple different geolocation chipsets. A Hadoop big data processing system was put in place to process the gigabytes...more of data collected from each test. This system cut the processing time of a single test from days to minutes. The insights from the collected data could then efficiently be viewed in almost real time by the test operator.
Entry level electrical engineer that assisted in the creation of telecommunications networks for different clients. Projects included the development of a mobile phone dispatch...more location in New Jersey, the plan for a local LTE network for Verizon Wireless, and performing field tests with different mobile equipment. All of these projects were developed using customized software for the telecommunications industry.
I built the entirety of the Automate Ads front (web) and back end systems. This includes all of the devops and data engineering work. Automate Ads product consists of a front end dashboard (backbone.js) connected to a backend REST API (Node.js). The front end dashboard allows customers to connect all of their paid ads accounts from Facebook, Google, Twitter, etc. and then collects the data from these accounts. From this data, users can then manage their ad campaigns, create new campaigns, and optimize cross channel. The largest portion of the product is the backend machine learning (neural network) system that has learned over time how to predict optimal bids and budgets for campaigns. This system uses this data to make decisions on the users behalf in order to optimize ad campaigns that would normally require substantial manual work. The current system is fully automated with 0 human interaction and saves customers over 20% of their budget of average.