- Strategically managed company-wide machine learning initiatives, aligning them with overarching business objectives.- Collaborated with primary stakeholders to pinpoint and leverage areas where machine learning offered a significant competitive edge.- Spearheaded technical proposals for high-value contracts and grants, emphasizing the development and implementation of novel machine-learning algorithms.- Directed the end-to-end research, development, and deployment of state-of-the-art ML/AI solutions, resulting in three production models.- Supervised the design and development of robust ETL/ELT and data labeling infrastructures using Python, SQL, and C++, which became the foundational platform for all ML projects.- Implemented and fine-tuned cutting-edge computer vision models for novel domains and applications.- Pioneered code optimization in C++ that achieved an 8x speedup, enabling real-time processing capabilities on edge devices.- Designed and integrated classification model into Kalman Filter for real-time target identification, association, and tracking using Python and C++.- Delivered a sequence of company-wide lectures and curated team-specific training sessions on machine learning and software engineering topics.- Received five peer-nominated cash awards from both my immediate team and inter-departmental groups in recognition of outstanding collaboration and innovative thinking.

Responsible for the software development and maintenance of tools used to test, diagnose, and verify embedded systems (e.g., radar and turret systems). Projects include:- Built Debian package to monitor application and hardware system health and debug issues.- Created Grafana monitoring interface with Loki and Prometheus backends.- Developed an automated end-to-end testing framework.- Updated C++ drivers for automated turret tracking system.

Developed methods in scientific machine learning and high-performance computing. Topics/projects include- Image segmentation tree-based domain decomposition for tabular data emulation- Gaussian Process Regression for tabular data emulation- Gaussian Process Regression / Radial Basis Function interpolation for improved High-Order finite volume scheme- GPU acceleration of hydrodynamics code

Developed open-source software to implement a novel tree-based lookup table emulator for physical tabular data in high-performance computing (HPC) environments. Aspects of the project include:- Parallel, thread-safe, and performant C++ code- Fortran and python bindings- Incorporating code into FlashX, an HPC astrophysics application- Implemented ideas from a paper I first authored that is currently under review

Development and optimization of FLASH's hydrodynamics solver. FLASH is a high-performance computing application used for astrophysical simulations. My projects include the following:• Using OpenMP 4.5/5.0 to offload computations to GPU• Parallelizing code with MPI• Scientific machines learning with Gaussian Processes

• Developed, Optimized, and debugged Python code to simulate a dynamic system using finite element methods and scientific python packages.• Enhanced existing computational simulations to use parallelization on Linux based supercomputing cluster, decreasing simulation run time by 50%.• Utilized Statistical and Mathematical data analysis techniques, including Fourier analysis, to interpret data.• Designed and created code to automate preprocessing, postprocessing, data analysis, and visualization, which allowed high simulation throughput and reduced operator's time to run each simulation by 99%.• Trained new team members and updated existing members on the development and use of simulation software through one on one mentoring and weekly group presentations.

• Developed a computational model using first principle physics to visualize and optimize physical systems using MATLAB.• Designed and drafted a proposal for a new product based on results from parameters optimized in simulations.• Presented research at APS conference and to researchers at Princeton Plasma Physics Lab.• Upgraded remote capabilities of a scientific experiment using LabView and C.