How Excel AI Agents Like Rosie Work for FP&A Tasks but Fail at Building Models with Giles and Ian

Note: This episode was recorded and scheduled for release before the recent announcement that Rosie AI is shutting down.

In this episode of the ModSquad series on Financial Modeler’s Corner, host Paul Barnhurst is joined by modeling experts Ian Schnoor and Giles Male to evaluate Rosie AI, a new tool that integrates with both Excel and Google Sheets. Together, they test its capabilities in building financial models, solving complex FP&A tasks, and performing real-world use cases. They push Rosie through a series of tests, from basic formula creation to building full three-statement models, and discuss where it excels, where it needs improvement, and its potential future in the world of financial modeling.

Expect to Learn

• How Rosie AI performs in real-world financial modeling tasks
• The strengths and weaknesses of Rosie in building models and formulas
• Key insights into how AI tools are evolving in financial modeling
• The importance of knowing how to validate AI-generated models
• How Rosie compares with other AI tools and traditional financial modeling techniques

Here are a few quotes from the episode:

• “AI can’t replace the need for modelers to understand the logic, but it can help speed up certain tasks.” - Giles Male
• “If you know what you’re doing, these tools can save you time. If you don’t, you need to double-check everything.” - Ian Schnoor

Follow Ian Schnoor:

LinkedIn - https://www.linkedin.com/in/ianschnoor/

Follow Giles Male:

LinkedIn - https://www.linkedin.com/in/giles-male-30643b15/

Rosie AI: https://www.askrosie.ai/

In today’s episode:

[02:23] - Excel Turns 40

[07:29] - Microsoft’s AI Agent vs. Modeling Tools

[16:18] - Exploring Rosie’s Website and Features

[25:35] - Real-World FP&A Test: Trial Balance Summary

[35:53] - Deferred Revenue Schedule Test

[50:54] - Detecting Inconsistent Formulas in Financial Models

[01:00:53] - Testing Rosie AI on a Five-Year Forecast Model Build

[01:08:31] - Final Thoughts: Where AI Helps and Where It Doesn’t