Sam Newman - Machine Learning Engineer
Sam Newman's walking deck for Machine Learning Engineering and AI Development
My Philosophy
My core philosophy centers on iterative innovation: Continuous refinement and data-driven learning are how to progress AI systems.
Experimentation as a Foundation: I embrace a culture of rigorous testing and controlled experiments to validate hypotheses and uncover optimal solutions.
Feedback as Fuel: I’m always seeking and integrating feedback loops - from users, data, and cross-functional partners - to constantly enhance performance and user alignment.
Adaptability as an Advantage: I recognize that the landscape of technology and user needs is ever-evolving, and maintain flexibility to pivot and optimize.
Featured Projects
Prompt Iteration Tool Demo - Babel Street
Objective
Made for an internal dev team to iterate on agent prompts, streamlining the development cycle for AI agents.
Approach
Developed rapidly to address an immediate team need for efficient prompt testing and refinement.
LSAT Practice Test Generator - Open Source
Objective
To generate practice tests with real LSAT questions from old tests using natural language prompts.
Approach
Developed a novel algorithm to create hyperdimensional contextual encodings of LSAT questions. This allows for sophisticated understanding and generation of relevant test content.
What People Say
Hear what colleagues and leaders have to say about my work.
“Sam is able to think deeply about problems and build solutions that align to organizational priorities and constraints. He is a strong coder, has a sound understanding of the AI and engineering technologies he uses, and has an uncanny ability to communicate clearly at any level of technical detail. On top of all this, Sam is generous, empathic, and genuine — an absolute joy to work with.”
— Brian Dalton, 2025
“There are few people that understand what it means to take accountability for their work, and that’s what Sam brings to the table. Not only does he have the technical and theoretical skills to carry out end to end ML projects, but he puts in the extra work to make sure that everyone is on the same page by communicating actively to make it happen. Having him on your team brings the sense of calm that only comes from having a dependable colleague.”
— Fiona Dixon, 2025
“Sam works skillfully with multiple stakeholders to build a solid understanding of each application of AI that can be built and evaluated with technical rigor, surfacing the multifaceted challenges of bias and ethics, assuring that our technologies deliver on their purpose. This “technical rigor” is no mean feat: task definition is the start, but Sam can take a project through the identification of data sources and the writing of annotation guidelines. Even though no one stakeholder (other than I, as his manager) can truly appreciate the importance of every dimension of his work, Sam maintains congenial relationships with all involved.”
— Jason Alonso, 2025
My Impact
My contributions at Babel Street and beyond have focused on accelerating development, improving communication, and fostering innovation.
Babel Street Impact
- Fast-paced development: Consistently delivered features ahead of schedule.
- Flexibility: Flexibly handled last-minute requests from product, ensuring project agility.
- Dev - Product Communication: Facilitated product team’s understanding of machine learning data pipelines and data considerations for quality and safety.
- Developer workflow: Regularly created tools to improve efficiency of team.
- Process documentation: Consistently pioneered and maintained detailed documentation to ensure any dev could understand my work and pick up where I left off.
- Policy: Was trusted by the VP of Machine Learning to spearhead the creation of AI policy, coordinating with product teams across the engineering organization.
My Skills & Tools
My expertise spans a wide range of Machine Learning and software development technologies.
Programming Languages
- Python (Advanced)
- JavaScript (Intermediate)
- Rust (Basic)
- Lua (Basic)
Machine Learning Frameworks & Libraries
- PyTorch
- Scikit-learn
- Pandas, NumPy, Polars
- Hugging Face Transformers
- Google GenAI API and Vertex API
MLOps & Cloud Platforms
- Git (Version Control)
- Docker
- Kubernetes (Basic)
- AWS
- Google Cloud Platform (Vertex AI - Intermediate)
- MLflow
Other Tools & Concepts
- Data Visualization (Matplotlib, Seaborn)
- API Creation (FastAPI, Flask)
- Agile Methodologies
- Technical Documentation
- Linux
- NeoVim
- SQL