Joe Gonzalez

I’m a Senior AI Engineer with over 6 years of experience building production machine learning systems, from model training and evaluation to the backend infrastructure that makes them reliable. I started at UCLA Health building HIPAA compliant data pipelines and APIs, then moved into machine learning at Newegg for search and personalization.

At Newegg, I developed and deployed deep learning ranking models that improved product search NDCG@10 by 8%, built end-to-end training pipelines with clickstream and user behavior signals using PySpark, and designed A/B experiments that lifted overall conversion rate by 2.3%. I also implemented a real-time feature serving layer that reduced feature computation latency from 200 milliseconds to under 20 milliseconds.

Now at Rho Technologies, I lead document intelligence initiatives across multiple languages. I train models like LayoutLMv3 on custom annotated datasets, architect document intelligence systems, and create evaluation frameworks and annotation systems that keep AI products trustworthy.

I work across the full stack of applied AI, closing the loop between training, rigorous testing, and real-world business impact—using automated evaluation harnesses, drift monitoring, closed-loop retraining workflows, and A/B experiments to improve extraction F1 and reduce manual QA and review load.