I’m an applied AI developer with 7+ years of hands-on machine learning experience, including senior-level work embedded within NASA’s High End Computing (HECC) program. I specialize in turning complex, messy data into reliable, scalable AI systems—not just demos. I work across: • Deep learning (PyTorch, TensorFlow, CNNs, RNNs/GRUs, Transformers) • End-to-end ML pipelines (from raw data to evaluation and deployment) • Large-scale & multi-GPU training • Model debugging, optimization, and performance tuning Before AI, I spent over 20 years in experimental science and earned a PhD in Chemistry, which means I approach ML problems with rigor, validation, and a strong understanding of real-world constraints. I’m especially effective when: • the problem isn’t fully defined yet, • the data is large, noisy, or scientific/technical, • or performance and correctness actually matter. Clients work with me as a senior technical partner, not just an implementer. I communicate clearly, move quickly, and focus on building AI that holds up outside the notebook. Let’s build something that actually works.

Thaddeus Norman

I’m an applied AI developer with 7+ years of hands-on machine learning experience, including senior-level work embedded within NASA’s High End Computing (HECC) program. I specialize in turning complex, messy data into reliable, scalable AI systems—not just demos. I work across: • Deep learning (PyTorch, TensorFlow, CNNs, RNNs/GRUs, Transformers) • End-to-end ML pipelines (from raw data to evaluation and deployment) • Large-scale & multi-GPU training • Model debugging, optimization, and performance tuning Before AI, I spent over 20 years in experimental science and earned a PhD in Chemistry, which means I approach ML problems with rigor, validation, and a strong understanding of real-world constraints. I’m especially effective when: • the problem isn’t fully defined yet, • the data is large, noisy, or scientific/technical, • or performance and correctness actually matter. Clients work with me as a senior technical partner, not just an implementer. I communicate clearly, move quickly, and focus on building AI that holds up outside the notebook. Let’s build something that actually works.

Available to hire

I’m an applied AI developer with 7+ years of hands-on machine learning experience, including senior-level work embedded within NASA’s High End Computing (HECC) program. I specialize in turning complex, messy data into reliable, scalable AI systems—not just demos.
I work across:
• Deep learning (PyTorch, TensorFlow, CNNs, RNNs/GRUs, Transformers)
• End-to-end ML pipelines (from raw data to evaluation and deployment)
• Large-scale & multi-GPU training
• Model debugging, optimization, and performance tuning
Before AI, I spent over 20 years in experimental science and earned a PhD in Chemistry, which means I approach ML problems with rigor, validation, and a strong understanding of real-world constraints.
I’m especially effective when:
• the problem isn’t fully defined yet,
• the data is large, noisy, or scientific/technical,
• or performance and correctness actually matter.
Clients work with me as a senior technical partner, not just an implementer. I communicate clearly, move quickly, and focus on building AI that holds up outside the notebook.
Let’s build something that actually works.

See more

Language

English
Fluent
French
Beginner

Work Experience

Senior Software Engineer at Intrinsyx Technologies Corp
April 1, 2018 - November 1, 2025
Staff data scientist at NASA’s High End Computing Capability (HECC); POC for all HECC user related data science and AI/ML issues, troubleshooting and resolving issues within 60 days; Maintainer of the HECC supported data science conda environments and miniconda, which was used by dozens of users per quarter who created hundreds of conda environments; Assisted in the testing and implementation of new data science related applications such as Dask, PyTorch Lightning, Podman, Ollama and Notebook Intelligence; Benchmarked single and multi-GPU compute node performance, including nodes with Nvidia A100 or GH100 GPUs, using scientific data sets; Leader of internal and HECC user sponsored data science projects; Insured information in the HECC Knowledge Base web-page for machine learning was current, correct, and actionable, including instructions on how use PyTorch and TensorFlow on compute nodes with multiple gpus and across multiple nodes with multiple gpus; NASA Hackathon Mentor; Acted as

Education

PhD at University of California, Santa Cruz
January 11, 2030 - January 1, 2004
Bachelor of Science at Santa Clara University
January 11, 2030 - January 1, 1995

Qualifications

Add your qualifications or awards here.

Industry Experience

Government, Software & Internet, Professional Services, Education