Open to opportunities

Sarah Bunks

@sarahbunks

Message

Machine Learning engineer specializing in NLP, transformer models, and production ML systems.

United States

Message

What I'm looking for

I seek roles where I can lead NLP model development and production deployment, improve model performance, and promote responsible, human-centric AI.

I am a Machine Learning Engineer with over a decade of hands-on experience building and productionizing NLP and ML systems. I blend deep technical expertise in transformer architectures with a human-centered approach to make AI both accurate and empathetic.

At H2O.ai and DataRobot I led integration of advanced NLP capabilities and optimized AutoML pipelines, reducing latency and training times while expanding text-analytics functionality. I also contributed to GPU-accelerated training at NVIDIA and research on few-shot learning and adversarial robustness at Stanford.

I enjoy mentoring, participating in Kaggle competitions, and supporting initiatives like Women Who Code, and I seek roles where I can drive model performance, production deployment, and responsible AI practices.

Experience

Work history, roles, and key accomplishments

Lead NLP/ML Engineer

H2O.ai

Jan 2019 - Dec 2023 (4 years 11 months)

Pioneered integration of advanced NLP capabilities into H2O.ai's Driverless AI, introduced transformer support and reduced deployment latency by 15% while enabling enterprise adoption through workshops.

NLP Transformers Python TensorFlow PyTorch Deployment Kubernetes Docker GCP)

Senior Machine Learning Engineer

DataRobot

Jan 2015 - Dec 2019 (4 years 11 months)

Led optimization of AutoML algorithms achieving 20% faster training times and integrated NLP capabilities to expand text-analytics use cases while contributing to MLOps tooling.

AutoML NLP Python scikit learn Optimization MLOps Jenkins AWS SQL

Research Assistant

Stanford University

Jan 2016 - Dec 2018 (2 years 11 months)

Collaborated with the Stanford NLP Group on multilingual NLP and developed a few-shot learning proof-of-concept and adversarial-robustness benchmark showcased at the AI symposium.

NLP Few Shot Learning Benchmarking Transformers Python Data Annotation Collaboration

Machine Learning Engineer

NVIDIA

Jan 2012 - Dec 2015 (3 years 11 months)

Optimized CUDA libraries and collaborated with hardware teams to fine-tune deep learning models for GPUs, improving performance by ~25% on targeted architectures.

CUDA Deep Learning Python PyTorch & Model Tuning Performance Engineering Research Collaboration

Education

Degrees, certifications, and relevant coursework

Stanford University

Master of Science, Computer Science (Machine Learning/AI)

2016 - 2018

Activities and societies: Research assistantship with Stanford NLP Group; developed few-shot learning proof-of-concept and benchmark dataset for transformer robustness.

Master of Science in Computer Science with a focus on Machine Learning and AI, including collaboration with the Stanford NLP Group on multilingual and few-shot learning research.