Post by Phacet

Last Wednesday, we brought together 97 finance professionals to talk about AI. The goal was to share the method we use at Phacet to practically transform finance operations, and the first agents teams should start with to avoid common pitfalls. It was also an opportunity to listen to our users, so we asked two simple questions. 1️⃣ What is your top priority use case? For 35% of the audience, it’s the work of cleaning, reconciling, and controlling data: essential for analysis and monthly close. 2️⃣ Where do you spend the most time today? Once again, at 32%: close & reporting. What’s the link between these two observations? The issue isn’t that the close is too slow. It’s that errors are discovered too late. To us, this is a key realization in finance: the greatest operational value comes first from data work, not from the ability to produce “AI-assisted” analysis. If the data isn’t reliable and properly reconciled: - analysis is not auditable - sources are unclear - and their overall value is limited The real promise of AI in finance, in my view, is to: - make data reliable before the close - detect variances as soon as they appear - reduce end-of-month stress - and ultimately make data truly usable (with or without AI, for that matter) 📊 The full results and detailed reasoning are available in the replay (contact us in PM if you're interested).