Artificial Intelligence and Machine Learning - C++-Created an extensible library implementation for Hidden Markov Models using the Baum-Welch and forward-backward algorithms.-Used LogSumExp techniques to manage underflow issues in the alpha/beta matrices-Used Eigen's tensor library and connected numerical analysis to a generic GPU object. -Implemented perceptron and multilayer perceptron with templated activation functions; Using TensorMaps and TensorRefs for a more DOP (data oriented programming) approach.-Implemented standard Kalman Filter using eigen tensors. I'm looking to implement an extended Kalman Filter next. C++ Concurrency, Low Latency, Futures, Asynchronous (github code available)-Implemented thread safe queue capable of servicing multiple producers and consumers.-The thread pool accepts generic tasks organized to run asynchronously.-The library can read large files (>50 million line inputs) concurrently, mutate the text, and write the results to disk efficiently.-Used doctest as part of the design-test-document cycle.Packaged implot into a cross platform dependency management and build system (build2) for C++ (github code is available). Implot is a plugin used to generate highly detailed graphs for a wide variety of scientific and financial applications.Working on a cache-aligned, lockfree, bipartite buffer (bip buffer / circular buffer) for embedded video and image processing.Python-Futures, Asynchronous Executors, and Asynchronous Program flow-Created a Priority Sorted Queue to handle high priority items in an MPMC program-Selenium in Python-Tesseract and PyTesseract-Unit Testing with Mocks, Spies, Fakes, Stubs-PyPy and JIT style compilations-Custom Wheel ModificationsAmazon AWS:Introduction to Using Lambdas, S3 buckets, IAM for data processing and thumbnail generation.

Used integral images (summed-area tables) to improve computational complexity of processing algorithms from quadratic to linear time. The implementation also cached computed results providing avenues for additional processing.Used test driven development (TDD) Doctest to provide confidence in algorithm results, error tolerances, and to ensure proper conversion between various number-spaces.Retrofitted project with cmake modules allowing tests discovery without explicitly adding cxx/hxx files to CMakeLists file. The project layout necessitated re-implementation and alteration of CMake modules to work with the build pipeline.Set up, ran, tested, and debugged docker images that processed geospatial data for trees points and tree tracts in a scaled server environment.Created a dev environment in RHEL 8 to match the build/runtime environment. Involved custom installation of RHEL to a WSL image, provisioning tool suites, and command line setup.Developed numerical techniques for the estimation of bathymetric depthmaps based on consistent shoreline and inconsistent deepest point information.The approach used data-oriented programming combined with sorted index buffers for speed.Used GDAL to manage projections and transformations of data between S2, WGS84, Flat-earth and other coordinate/addressing systems.

Restructured user workflow in QT client application with input from sales team using mockups. Utilized Type Erasure and std::variant to handle data combined with an interface to QT's signaling system to allow reuse of GUI elements between both QDialog and QWizard.Expanded data polling capabilities between networked audio devices. This included SSH'ing into devices, attaching gdb (via Eclipse), debugging both the audio player and the audio amplifier.Wrote guide on how to setup ARM DS5 with Eclipse (embedded) IDE project to work on RidePlayer 25H network audio cards.Validated network functionality of 25H audio cards by writing embedded unit tests in cpputest. Expanded and restructured other tests so that they would expand in response to changes in production. Eg: I made one struct drive both testing and production. When the structure changed, tests would look for new sizes and data without necessitating a rewrite. The user could program against the new struct immediately.Utilized Enterprise Architect to generate mockups and mindmaps to understand customer needs with the help of the sales team.

Upgraded Auto Align Software Suite to dynamically interface and control multiple displays as a foundation for 120 hz and multiple-capability displays (daytime and IR displayed together).Analyzed graphics pipeline created a targeted plan to implement Quaternions to reduce memory footprint and speed up rendering.Vulkan research and familiarization with graphics pipeline (comparing concurrency and control with OpenGL), build pipeline, and validation layers. Part Time Scrum Master for a large design effort.Utilized standard Atlassian tools suite during development (Jira, Crucible, Bamboo/Jenkins, etc).Actively worked to incoporporate and document Design Patterns including Builder, Facade, Observer, Visitor, Adapter, and Abstract Factory using Curiously Recurring Template Pattern for static poly-morphism. Proactively learned and used Enterprise Architect to prototype code and keep UML diagrams updated with the latest prototype changes. The goal being to improve my personal code quality and processes.Wrote a custom container from scratch to mimic functionality of std::vector with the additional ability to change reallocation sizes as desired. This included custom iterator and allocator classes allowing it to work with std algorithms and opereations.Created a python SVN Merger that allowed users to merge repos made of SVN Externals without having to merge each external manually.Implemented Asynchronous GRPC streaming capabilities.

Experience with both HLA and DIS simulation.Implemented HLA interactions tunneled inside of DIS pdu’s, read in inputs, handled and parsed messages, and routed message contents to appropriate receiving functions.Developed, deployed, documented, maintained, and provided support for a virtualization and used by around 40 people in the department and an unknown number facility wide. The virtualization reduced the spin-up time for configuring the environment from 3+ months to a few hours.Scrum master for a major development effortLed peer reviews for the creation of new product featuresMapped and analyzed current processes, sought and built consensus, implemented changes to improve both software practices and processes.In-depth experience with Archlinux, Linux Mint, SLES, Fedora, Red HatEnterprise, Vagrant, Ansible, Python, Ruby, Perl, MySQL, Nginx, Passenger, Apache, Tomcat, REST and WADL, Qt, GTK/Glade, and GITFamiliar with Atlassian suite (Confluence, JIRA, and Bitbucket) as part of an Agile-style workflow

Responsible for daily testing, operational maintenance, and troubleshooting, and enhancement of four advanced development labs during second shift.Worked to commonize Gulfstream’s visual display architecture between host simulation and both FSI Vital and X-Plane targets by improving UDP setup, logical devices, model/plugin development to improve API usage, encoder/decoder libraries, and UML documentation using OOA.Improved Model Verification and Validation processes by writing tools to setup, run checks on Simulink models, and then export those results to report files.Utilized XML 1.0 and Schematron along with the Xerces (Java) parser to both validate and create MATLAB objects on the fly. The resulting architecture was extremely extensible and supported significant changes without rewriting core code.Worked to commonize navigation databases in Host and Visual simulations by using ARINC 424 compliant database sources.Utilized concepts and tools like MATLAB/Simulink, Java, Python, ADI, Distributed Simulation and Real-Time Simulation Techniques (both UDP and reflective memory based), C/C++, UNIX/Linux, SVN, and more.Helped integrate, setup, and test LRUs and simulation chassis from Speedgoat and ADI within a real-time simulation environment for multiple labs.Modified and integrated C/C++ drivers for A429 cards used by Speedgoat so that simulation could transmit a JSON-like data stream formonitoring by Lab equipment (LabVIEW).Worked with Field Programmable Gate Arrays (FPGA).Presented an Introduction to Software Configuration Management based heavily on IEEE Standard 828-2012.

Developed in-depth real-time training environments for Chinook andBlackhawk using advanced software techniques coupled with Agile stylemethodology.Set up sound simulation adding ASTi (Telestra) to distributed system. This required setting up multicasting, using tcpdump for troubleshooting, and general network configuration. I additionally created a communication class in the Host simulation to manage the sound simulation via UDP.Was personally responsible for Chinook Desktop Trainers for avionicsversions 8.4, and 9.2, as well as Electrical Fault, and ElectroMechanicalFault Trainers.Used Tools and languages including Visual Studio, Cygwin, Linux VirtualMachines, C++, Python, Bash, GLSL/HLSL shaders, ASTi (Telestra), and OpenscenegraphImplemented OpenSceneGraph multitexturing by using GLSL shaders and MFC multitexturing by using double buffering and alpha blending.Mini-projects included implementation of DAFIF database withSIDs/STARs/IAPs flight plan insertion and modification, creating andenhancing virtual MFD’s in OpenGL, and integrating faults and malfunctionsfor training purposes into the environment according to customerrequirementsDid a large amount of work organizing projects in Project Managementsoftware as well as doing three man-hour estimates for new contracts orenhancements. Wrote supporting documentation for many new processesthat I developed.Configured Virtual Machine (CentOS) with NetBeans and SVN and paired with a user setup shell script that automatically setup users and their credentials. This time saving measure allowed a quick project ramp-up.

Developed graphic displays (in C++ VS2010) and logic (in C++ and Fortran VS2010) for MFDs, PFDs, Sensor Displays Units, and Radio Tuning Units in aircraft simulation on Desktop Training SystemsHave done development work to produce interfaces for previously unsupported navigation equipment including porting a TCAS suite from Fortran90/C++ in Windows to Fortran75/C in a Unix Environment. This was to update Traffic and Collision Avoidance across the older simulators in the fleet.Simulated systems included GPS, TCAS, TAWS+, AHRS, IRS, Radio Altimeter, FMS, Adaptive Flight Displays, EGPWS, MLS, ILS, Comms Radios, WXR input and Terrain inputs.Developed Malfunctions for navigation equipment that simulate failing equipment to better train clients.Participated in projects including Bombardier Challenger 605, Hawker 750XP, Hawker 900XP, Gulfstream Galaxy 200, Embraer Dash 8 (TCAS port), Cessna CJ4, BeechAir C90GTi, Hawker 400XP, BeechAir Premier 1A, Embraer 170/190 ConvertibleVery Familiar with ARINC429 Standards, Honeywell Epic systems, Proline 21, Proline 4. Have experience interfacing with both rehosted software suites and physical hardware in full flight simulators

Investigated the reliability of sEMG and MMG intensity patterns of the leg extensor muscles produced by wavelet analysisUtilized National Instruments consoles to gather electronic signals for processing Wrote unique MatLab program to analyze sEMG and MMG Data using nonlinearly scaled wavelet analysis and principal component analysisWrote unique LabView programs to collect sEMG and MMG signal data from the National Instruments Console and peripheral equipmentHandled digital signal processing of all data using Matlab and Labview

WinRobo is a design competition where student groups design a robot weighing less than 1kg capable of cleaning a double-hung sash window. Designed large part of the electronics package that controls the window cleaning robotSelected sensor package and arrangement and wrote C programs to allow an embedded microprocessor to interface with the packageWrote a unique embedded real-time program in C to control and allow the WinRobo to automatically adjust to differing sized windows, cross over to the upper window pane and continue cleaning

Tutored disadvantaged, first-generation, and culturally diverse student populationRequired knowledge in Math, Science and Engineering courses relevant to the engineering curriculum

Tutored Naval Officer candidates in their required study of Calculus and Physics and science-based subjectsSecondary mission was to provide study help in a variety of engineering courses and other science and math coursesSubmitted student usage and progress reports as part of an effort to track the students’ advances

Installation, operation and performance of unit level maintenance on multi- functional/multi-user information processing systems, peripheral equipment, and associated devicesPerform analyst functions; construct and test computer system programsMaintenance, troubleshooting, and installation of the Reserve Center’s intranet.

Investigated the location of the innervation zone of the Vastus Medialis as a function of height and femur length Gained in-depth knowledge of National Instruments consoles by wiring connections for mechanomyograms (MMG)Used a unique 16-element linear array of surface electromyogram (sEMG) sensors during investigationHandled digital signal processing of all data using a combination of Matlab, Labview, and Excel (to display data)

The Oklahoma Department of Human Services sponsored Kid’s Café as a program to provide daycare to underprivileged children. As Summer Session Coordinator, detailed monthly reports were submitted to DHS that tracked the general progress of children in the program, the exact number and ingredients of every meal served, as well as individual success stories.Implemented a flexible time management system to manage the childrenCoordinated with local businesses for additional funding as well as field trips