Welcome to my corner of the internet! I'm an innovative and scientifically rigorous Machine Learning Engineer with expertise spanning applied research, production-grade engineering, and MLOps infrastructure.

I'm adept at transforming complex theoretical concepts into scalable, real-world AI/ML systems, from initial experimentation to containerized deployment and monitoring in mission-critical environments. I specialize in Computer Vision, Space Domain Awareness (SDA), and advanced analytical algorithms.

Machine Learning Engineer at Turion Space

Irvine, CA • June 2024 – Present

• Leading end-to-end development of 3D reconstruction algorithms for space RPO using ViT architectures and 3DGS
• Developed multi-model RSO characterization pipeline using PyTorch, integrating CNNs, Unet, and YOLO architectures
• Built production-ready APIs serving containerized model pipelines using Docker and KServe on Kubernetes
• Automated 3D synthetic data generation in Blender and deployed active-learning data-labeling workflows

Machine Learning Researcher at UC Merced

Applied Math Department • Merced, CA • June 2023 – June 2024

• Co-authored research using physically informed CNN (PINN) for high resolution synthetic aperture radar (SAR) imaging
• Optimized CNN using TensorFlow to overcome convergence issues, achieving 88% prediction accuracy
• Implemented physically informed loss function, increasing recovery rate by 8% while reducing model hallucinations
• Developed custom metrics for extended scatterers, providing more precise assessment of predictions
• Successfully translated and debugged 500+ lines of code with extensive testing for model robustness

Technical Skills

Languages

Python (fluent), SQL (proficient), C++ (intermediate)

Technologies

PyTorch, 2D/3D Vision, LLMs, NumPy, NVIDIA Inference Server, BlenderProc

Cloud & Infrastructure

Kubeflow Pipelines, KServe, Kubernetes, Docker, CI/CD, AWS, Azure