Tiffany Ma, Director of Product Marketing for AI and Operational Analytics at OneStream, returns for her second podcast appearance to share what’s really happening with AI in finance. From 100+ live customer implementations to natural resource forecasting, the AI hype is meeting actual results.

Remember when Tom Shea showed IBM Watson back in 2017 and shocked everyone by saying “this is really hard, I’m not gonna sell it to you”? That honesty was refreshing, but also showed the problem – AI was expensive, complicated, and required data scientists writing Python for years. Fast forward to today and OneStream’s cracked the code on what they call “applied AI” – purpose-built for finance, no coding required, delivering results in weeks instead of years.

But here’s what’s really interesting – the use cases have exploded beyond what anyone expected. Started with demand forecasting, expanded to P&L, now customers are doing workforce planning, cash management, even trying to forecast the materials needed to produce steel. That’s not your typical FP&A use case, but when you have a tool that’s agile enough to iterate quickly, people start getting creative.

Tiffany breaks down why OneStream renamed Sensible Machine Learning to Sensible AI. Machine learning is quantitative – numbers forecasting numbers. But they’ve added generative AI for automated commentary and variance analysis, plus agentic AI with their new agents. They couldn’t box themselves into “machine learning only” when the solution does way more than that.