← Back to Speaking

Beyond Compliance Theater: Building Production-Ready AI Governance with Multi-Agent Workflows

Event: APIDays London 2025
Location: London, UK
Date: September 22, 2025

In this talk, I explored the foundations of what makes an AI agent an agent, introduced the FINOS AI Governance framework and showed how to implement some foundational controls using open source software

Key Topics Covered

What Makes An Agent an Agent?

  • Agents vs ChatBots
  • Agent Architecture 101

An Intro To AI Governance

  • 90% the same, 10% different and what that means for you
  • FINOS AI Governance Framework - what is it and why should you care?
  • Why Multi-Agent makes things much worse
  • Balancing data quality with system performance

Enterprise Patterns for Agents

  • How to classify AI risks effectively
  • An enterprise agent architecture
  • Practical deployment examples for Kubernetes Gateway API

Audience Takeaways

Attendees gained a grounding in agentic AI architectures, financial services specific AI governance frameworks, and how to mitigate some of these risks in open source software.