Your enterprise AI agents are up and running. The outputs look polished. The answers sound confident.

So why do decisions still feel risky?

Richard Sonnenblick, Chief Data Scientist at Planview, has built decision systems in high risk industries where bad data leads to real consequences. He argues the biggest challenge with enterprise AI is not autonomy. It is data quality and human ownership.

In this episode, Rich explains how outdated systems of record quietly train AI on optimistic fiction, why garbage in still means garbage out, and what practical agent driven workflows can prove ROI in the first 60 to 90 days.

If you are scaling enterprise AI across your teams, this conversation will change how you think about accountability and execution.