Austin, Texas, United States
I am a results-oriented engineering scientist and software engineer with extensive expertise in geospatial deep learning, computer vision, and high-performance computing. My expertise focuses on the development and optimization of feature extraction methods, point cloud processing, video compression techniques, and machine learning applications. With a solid technical background in C++ and Python I create and implement high-performance, scalable software solutions that push the limits of computational efficiency. My work focuses on addressing hard technological issues in AI-driven analytics, image processing, and scientific computing. I have strong expertise with parallel computing, DevOps CI/CD pipelines, and cloud computing (AWS/GCP), which enables me to build resilient, scalable systems. Beyond implementation, I actively contribute to patent creation and research-driven innovation, ensuring that my work has a long-term influence in the sector. I am enthusiastic about mentoring and technical leadership, and I like guiding research teams, younger engineers, and graduate students through cutting-edge software development, algorithm engineering, and best practices. Whether I'm improving deep learning models, speeding GPU-based computations, or designing next-generation geospatial analytic tools, I thrive in places where innovation meets practical application.
Lead the development and deployment of geospatial deep learning classification and feature extraction systems to accurately identify and classify objects such as trees, vehicles, and buildings from satellite-derived 3D datasets. Spearhead enhancements to NASA’s Sliderule Earth project, improving cloud-based geospatial analytics for on-demand, high-resolution data delivery and optimizing subaqueous terrain depth estimation using deep residual networks and machine learning models. Design and optimize a high-performance orthometric image processing system, reducing classification error rates by 20% and accelerating processing speeds through C++ optimization, parallel computing, and advanced 3D rendering techniques. Develop and refine change detection algorithms for large-scale point cloud datasets, enabling automated damage assessment and anomaly detection for time-series 3D geospatial analysis.
Designed, developed, and optimized advanced video compression algorithms, achieving higher encoding efficiency and superior image quality while integrating into industry-standard frameworks such as FFmpeg, h264, and h265. Built and maintained CI/CD pipelines and DevOps workflows, automating software deployment, version control, and testing for scalable high-performance video processing systems. Secured two patents in video compression and image quality assessment, contributing groundbreaking innovations to video encoding and streaming technologies. Engineered and optimized high-performance software solutions for Linux-based platforms, leveraging C++, Python, and GitHub for efficient version control, parallel computing, and multi-threaded processing.
Led research and development in geospatial deep learning classification, creating novel machine learning algorithms that improved classification speeds by 60× and reduced error rates by 50% in primary geospatial detection tasks. Served as Principal Investigator for a large-scale geospatial feature extraction modernization project, implementing DevOps workflows, CI/CD pipelines, and cloud-based computational frameworks to enhance scalability and automation. Acted as a technical liaison and advisor to the National Geospatial Intelligence Agency, contributing to backend integration efforts for large-scale geospatial data services and classification models. Designed and implemented high-performance computing benchmarks for feature extraction algorithms, improving computational efficiency in sonar imaging, geospatial analysis, and 3D reconstruction techniques.
Conducted applied research on image denoising and super-resolution, developing statistical modeling techniques and advanced filtering algorithms for solving ill-posed problems in point prediction. Applied high-performance computing (HPC) techniques using the Texas Advanced Computing Center’s (TACC) peta-scale compute cluster, enabling large-scale image analysis and data modeling. Designed and implemented machine learning-based image processing algorithms, leading to two awarded patents in point prediction and image denoising techniques for digital imaging. Provided mentorship to graduate students and researchers, offering guidance in experimental procedures, software design best practices, and AI-based image processing techniques.
Led a web development team in designing, developing, and deploying an interactive research application site for natural image processing, ensuring scalability, high-performance rendering, and real-time data processing. Designed and conducted experimental research in computational imaging and vision systems, analyzing human perception and its implications in computer vision applications. Developed and implemented psychovisual experiments using custom-built software, enabling real-time adaptive gaze-contingent visual field simulations for neuroscience and human perception studies. Provided mentorship to graduate students and research associates, fostering technical expertise in software engineering, programming best practices, and AI-driven vision research.
Designed and developed a real-time software imaging system based on a human foveal vision model, integrating advanced real-time simulation techniques for high-fidelity image processing. Created the XGL software library, an advanced real-time visual experiment framework, widely adopted in university labs and research institutions for computational vision research. Acted as a technical liaison and research collaborator for global institutions, including MIT, University of Chicago, FEEIT Skopje, Sapienza University of Rome, and NYU, supporting XGL software applications in vision science. Led the planning, implementation, and performance benchmarking of high-performance computing solutions in visual computing and real-time image processing applications.