Open to opportunities

David Bourgin

@davidbourgin

Message

Senior Machine Learning Engineer specializing in scalable real-time AI detection and anomaly systems.

United States

Message

What I'm looking for

I seek roles building and scaling production ML systems for real-time detection and embeddings, in collaborative teams with strong MLOps practices and impact-driven missions.

I am a Senior Machine Learning Engineer with 14+ years building scalable AI detection systems across startup and enterprise environments. I specialize in malicious content detection, real-time inference, embeddings, and distributed data processing.

At Adobe I architected real-time inference pipelines processing 500M+ daily events, achieving 99.2% detection accuracy and reducing false positives by 35%. I also scaled similarity-search and embedding systems to handle 10B+ domain records while improving precision and lowering computational costs.

Previously at Spotify I built recommendation and batch inference pipelines processing tens of billions of events, improving engagement and reducing latency dramatically while preventing multimillion-dollar fraud loss. My academic work includes doctoral research and postdoctoral work on interpretability, adversarial robustness, and anomaly detection with multiple peer-reviewed publications.

I focus on delivering production-ready ML solutions that adapt to adversarial threats, mentoring engineers, establishing MLOps best practices, and driving cross-functional collaboration to shorten model-to-deployment cycles.

Experience

Work history, roles, and key accomplishments

Current

Senior Machine Learning Engineer

Current

Adobe

Feb 2022 - Present (3 years 11 months)

Architected real-time inference systems for malicious content detection processing 500M+ daily events with 99.2% accuracy and reduced false positives by 35%, while cutting model-to-deployment time from 6 weeks to 2 weeks.

PyTorch TensorFlow Apache Spark AWS Sagemaker Embeddings Similarity Search MLOps

Machine Learning Engineer

Spotify

Jan 2020 - Feb 2022 (2 years 1 month)

Designed and deployed real-time recommendation models using embeddings and similarity search that improved user engagement 18% for 150M+ users and reduced recommendation latency from 800ms to 120ms.

Postdoctoral Researcher

Princeton University

Aug 2018 - Nov 2018 (3 months)

Conducted research on neural network interpretability and adversarial robustness, developing anomaly detection techniques that achieved 94% detection accuracy on benchmark datasets.

Anomaly Detection Deep Learning Model Evaluation Research Methodologies

Machine Learning Researcher

University of California, Berkeley

Aug 2012 - Nov 2018 (6 years 3 months)

Conducted doctoral research on deep learning and NLP, publishing 8+ peer-reviewed papers and developing scalable classification and anomaly detection algorithms for large-scale datasets.

Deep Learning Natural Language Processing PyTorch TensorFlow Distributed Training Hadoop Apache Spark Research

Research Fellow

Microsoft

May 2017 - Aug 2017 (3 months)

Developed ML models for large-scale content moderation and threat detection processing 100M+ daily data points and improved moderation latency by 25% via similarity search and ranking optimizations.

Content Moderation Threat Detection Embeddings Similarity Search Distributed Computing

Education

Degrees, certifications, and relevant coursework

University of California, Berkeley

Doctor of Philosophy, Computer Science / Machine Learning

2012 - 2020

Activities and societies: Published multiple peer-reviewed papers; designed distributed data processing experiments using Hadoop and Spark; mentored graduate and undergraduate students; collaborated with industry partners on applied ML projects.

Completed doctoral research in deep learning and natural language processing with publications on scalable model architectures and real-time inference systems.