Speaker: Carles Ferrater, Solutions Engineer (MENA), AIRIA

Most AI demos look impressive—until real users show up. In production, you need agents that are reliable, traceable, secure, and integrated with real workflows.

In this session, Carles will walk you through how to take an AI agent from a quick prototype to a production-ready system using AIRIA—covering the practical architecture, key components, and the pitfalls teams typically hit when they scale beyond a proof-of-concept.

What you’ll learn:

• What “production-ready” really means for AI agents (beyond a working demo)
• How to design agent workflows that are consistent, governed, and maintainable
• Integrating tools/APIs and enterprise data safely (with guardrails)
• Observability basics: logging, tracing, and monitoring agent performance
• Evaluation mindset: how to measure quality and reduce failures over time
• A clear roadmap to ship agents that businesses can actually trust

Part of the AI Building Application Challenge

Who should attend: builders, engineers, product teams, and AI leaders who want to move from experiments to real deployments.