Kyle Bailey

I am a Senior AI and Machine Learning Engineer with 12+ years designing, launching, and operating large-scale production AI systems, focused on LLM reliability, evaluation, and inference behavior under real-world traffic.

I've been technical owner of end-to-end LLM platforms, including retrieval pipelines, online and offline evaluation, observability, and incident response, supporting many product teams and external customers.

My work reduced P95 latency significantly, scaled retrieval-augmented generation systems to millions of documents and hundreds of thousands of daily queries, and introduced traffic-derived evaluation, edge-case mining, and online shadow testing to detect real-world failures.

I pair strong hands-on engineering with data engineering and analytics expertise—building dbt analytics layers, Spark/ Airflow pipelines, and human-in-the-loop feedback loops—while mentoring engineers on production LLM debugging, evaluation methodology, and reliability best practices.