Michael Mahon

I am a data scientist and machine learning engineer with 6+ years building LLM-powered, generative AI and predictive systems at Meta, Amazon, and startups. I specialize in turning research-quality models into production-grade services that drive measurable business outcomes.

I have led projects that produced a $200M uplift in customer spend at Amazon, reduced patient readmissions by 52% at SAIVA AI, and scaled Meta’s generative assistant to 400M+ monthly users while integrating LLaMA 3.2 multimodal capabilities. I optimized inference pipelines to cut latency and serving costs, and designed large-scale A/B experiments that delivered consistent engagement lifts.

I focus on robust MLOps, experimentation, and scalable deployment using PyTorch, TensorFlow, MLflow, Docker, and Kubernetes, and I enjoy mentoring cross-functional teams to align AI systems with product goals. I seek roles where I can build reliable, impactful AI systems and lead technical execution from research through production.