• Promoted within 13 months of start date due to exceptional project quality and organizational impact.• Identified and remedied scalability issues with data storage and processing procedures. Optimized processing 100 billion + terabytes of information by remapping data schema to a compact form. Required data storage was reduced by 90+%, and processing time by 85+%.• Directed the code development and data ecosystem of critical projects using Apache Hadoop and Apache NiFi for Extract-Transform-Load (ETL). Reduced the number of Jira tickets by 63% across projects.• Facilitated communication between developers and non-technical stakeholders between government agencies to ensure mutual understanding of project goals and communicate the needs of all parties. Teams completed 28% more Jira tickets per Agile sprint.• Standardized multiple Jira projects by overhauling ticket workflow paths and Kanban board structures, resulting in efficient ticket progression and template adoption across the broader office.

• Developed and deployed automated reporting and filtering algorithms to routinely audit critical data input sources for duplication and tracking spikes and drops in volume. This product found a 63% overlap across data sources and reduced the amount of data to process by 45%.• Built web-scraping utility to capture various cyber-threat intelligence information. This tool reduces manual research time by 1100 hours per year, per person, and automatically updates internal records of attack vectors.• Developed a flowchart generator to combine system configuration files into a visual overview graph using GraphViz, and provides data specification metrics. Manual documentation time was reduced by 2900 hours per year, per person.• Created an algorithm to classify computer network hosts based on behavior patterns observed in traffic, reducing analyst research time by 700 –1450 hours per year, per person.

Anomaly Detection through Machine Learning:Created a cyber-physical monitoring system using camera, sensors (light, sound, motion, etc.), Arduino Uno and Raspberry Pi along with machine learning algorithms to build a predictive model describing the environment’s state. Proof-of-concept prototype is built, functionally accurate, and will scale with additional sensors. Work was featured in poster presentations at the RSA Conference 2019.

Established an environment of virtual machines for students using existing cluster of servers and the VMWare ESXi hypervisor operating system. Productivity was greatly boosted with low overhead, a scalable infrastructure and no additional cost.

Mentored students entering the early college program at NC A&T State University. Students were tasked with year-long research projects and mentors ensured consistent progress. Managed project timelines, duty delegation, and teaching literature review and report summaries for the students. Students received great reviews on their final presentation and documentations.

Trained multiple classrooms of 35+ students in Java programming. Subjects included class inheritance, encapsulation, polymorphism, abstraction, GUI programming, file I/O, secure programming, event handling, data structures and algorithm design. Examination grades increased by 1-2 grade levels per semester.

Educated new students on standard research principles including literature review, source citations, accurate sourcing, paper formatting as well as the machine learning concepts that will be needed to conduct their work. This effort resulted in very positive reviews of their final poster presentation and gave them a much deeper understanding of the applications of mathematics pertaining to data science and machine learning.

Built a content migration web application using AngularJS, PHP, HTML/CSS, JavaScript and Windows Server 2016 management tools. Application was installed to the root directory of target web server and displays all files and directories to the user. The user can specify files and file types for transfer, then choose either a local destination or external when supplying an IP address and FTP credentials. Company now has an easy-to-use tool that streamlines content migration from server to server.

Developed web application to track and display UAV flight patterns from KML files using AngularJS, JavaServer Faces (J2EE), Tomcat Server, HTML5/CSS3 and Google earth’s API. This resulted in 12.5% faster research time, and allows non-technical users to draw insight from data.