Engineering AI That Actually Works

Written By joel benedict

In this episode of Pittsburgh Tech Council’s TechVibe podcast, Bill Elm from RCS explains the hype around AI and why 92% of corporate AI projects fail to meet business objectives, and how AI engineering—not experimentation—can prevent costly failures. Bill explains how RCS’s analytic engineering process, Brittalytics, exposes hidden flaws that can cause these failures and how to avoid costly or dangerous system breakdowns.

Listeners will learn:

  • Why most AI initiatives fail before they start—and how to fix that.

  • The difference between AI research and AI engineering.

  • How RCS’s “Brittalytics” process exposes hidden flaws that can cause costly or dangerous system breakdowns.

  • A real-world case study showing how RCS turned a rejected AI system into a high-performance solution embraced by users.

  • Why understanding human decision-making is key to making AI work in “a messy world.”

Learn more about Brittalytics and how to make your AI work with you:
Brittlaytics Audit Service
Brittlaytics Audit Tool

joel benedict
Next

RCS Developed the Brittleness Audit in Supporting Laboratory of Analytic Sciences (LAS)