I was invited to speak at the Generative AI Summit London, held at the InterContinental at The O2 in December 2025. My talk, "The hybrid pipeline: orchestrating fine-tuned LLM classifiers with science-based GenAI behavioral feedback at scale", showed how production AI systems layer fine-tuned LLM classification, GenAI classification and generation, and deterministic scoring, and when to reach for each: response time, inference cost, accuracy, data privacy and user experience.
The deep dive walked through the hybrid pipelines behind the leadership skills platform now called Hasha, with a live demo on stage, and closed with hybrid patterns for content moderation and customer service systems.