I recently completed a new portfolio project, the AI CFO Challenge: Evaluating AI Financial Analysis and Business Decision-Making. For this project, I wanted to see how well AI could build and analyze financial information, while also giving me an opportunity to use my Excel skills and business experience to evaluate the results. I used two different AI models. The first model generated budget and actual financials for a fictional 500-person advertising company. I moved the results into Excel, checked the calculations, and identified several errors in the AI-generated financials. After correcting the results, I also created a simple financial dashboard. I then provided the corrected financials to a second AI model and asked it to independently identify the five most significant issues and provide recommendations. I intentionally did not provide my dashboard or conclusions, as I wanted to see what the model would identify on its own. I evaluated the recommendations using a rubric across five areas: • Financial Accuracy • Business Logic • Risk Awareness • Business Applicability • Strategic Value What I enjoyed most about the project was combining Excel, budget planning, financial analysis, and my 30+ years of business experience with the AI evaluation and prompt engineering skills I have been developing. One of my biggest takeaways was that AI generated the financial information and analysis very quickly, but I spent considerably more time checking the calculations, reviewing the financial relationships, and evaluating whether the recommendations made sense from a business perspective. For me, this reinforced the importance of not accepting an AI response at face value, particularly when the results may be used to support a business decision. When using AI for financial or business analysis, where do you think human review adds the most value?