Building Production-Grade Agentic AI

**PRESENTATION OVERVIEW**

Building an AI demo is easy. Shipping an agentic system that thousands of real users depend on daily is a completely different problem. Roberto Moreno has done it — multiple times — at PLACE, including Ask Gabbi (an AI-first assistive OS inside a CRM), AI call coaching, and live lead qualification agents. This talk is the unfiltered technical and product reality of taking agents from prototype to production: reliability, failure modes, human-in-the-loop design, trust layers, and what "done" actually looks like. Key Takeaways: • The 3 failure modes that kill production agentic systems (and how to design around them) • Human-in-the-loop: when to require it, when to skip it, and how to build trust over time • Multi-agent orchestration in real workflows — what the architecture actually looks like • How to define "production-ready" for AI systems with no clear spec

**Speaker**:

[Roberto Moreno](https://www.linkedin.com/in/robertoamoreno/)

**Event Details**:

* L