
AI Today Podcast Why CPMAI Matters in AI Projects — with Mike Hyzy
27 snips
Feb 18, 2026 Michael Hyzy, VP of AI Strategy and Product Development at CGI, brings enterprise AI and governance experience. He discusses why treating AI like traditional software leads to stalled pilots. He explains CPMAI’s role in taming data uncertainty, shifting project manager mindsets, moving pilots into production, and preparing organizations for a future of AI agents and hybrid workforces.
AI Snips
Chapters
Transcript
Episode notes
Failure Led To Method Change
- Michael Hyzy describes a failed 18-month ML project that fell apart when treated like traditional software development.
- Adopting CPMAI refocused the team on data understanding and recovered lost time by structuring iterative hypothesis testing.
Stop And Fix Data First
- Pause model building and perform thorough data understanding and preparation before chasing metrics.
- Treat phases as variable-length and let phase three take as long as needed to avoid rework.
Embed Governance Early
- Build governance, bias testing, and ethical review into AI projects from the start rather than retrofitting them.
- Use CPMAI's phase discipline to satisfy regulated-industry requirements and maintain auditability.
