Speaker
Abstract
Most AI projects impress in demos but fail in production. While 60% of CIOs expect GenAI workloads by 2025, only a few deliver meaningful results. As a forward-deployed engineer implementing agentic and analytics systems in complex enterprises, I’ve seen why: reliable AI depends less on models and more on people, process, and disciplined delivery.This talk distills lessons from real-world deployments—how to build context within model limits, set up evaluation frameworks, structure delivery teams, and establish feedback loops that make AI smarter over time. Attendees will leave with practical patterns for engineering AI systems that stay reliable long after the first deployment.
Sponsored session
Session sponsored by
PromptQL delivers an “AI analyst” — one that is specialized for your business, reasons like your best analyst, gives trusted answers instantly, and continually learns and improves.