Farati Okiuvo

At Netflix, I lead development of a cloud-native multi-agent personalization platform on AWS, powering large-scale recommendation systems with streaming data, distributed feature pipelines, and low-latency inference.

I build Rust and Go microservices, Python AI orchestration, embedding retrieval, feature stores, and training-serving consistency frameworks that improve personalization quality, reliability, and serving efficiency. I focus on performant systems that remain useful, explainable, and trusted by the teams who depend on them.

Previously at Guidewire, I delivered an event-driven risk and fraud decisioning platform using Java, Spring Boot, Python, Kafka, Spark, Oracle, and PostgreSQL. At Dolby, I built cloud-native media processing services on Google Cloud that scaled to millions of events per day, reduced processing latency by 45%, and improved throughput threefold.

My earlier work at EY and Itree included modernizing Java applications, building secure REST APIs, optimizing financial data processing, and strengthening CI/CD and event-driven backend foundations.