Webinar
Join Numeric and Nicolas Boucher, the #1 AI in Finance influencer, who has trained 50,000+ finance pros at companies including KPMG, Mercedes-Benz and Siemens.
Every finance team is being told to build AI agents. Very few are told what to build them on.
An agent is only as good as what it can see. A journal entry holds an account, a debit, a credit and a short memo. It says what moved. It almost never says why. So AI fills the gaps with guesses, and you spend your hours checking its work.
The teams getting agents to work haven't always found smarter tools. They've often fixed the data. In this masterclass you'll watch how to fix this step by step, with practical use cases you can use as soon as the session ends.
You'll leave knowing what your data needs before AI can be trusted with it, and what to do first this month.
The 3 checks to run before you hand any file to AI. Skip them and you'll end up trusting an answer that's wrong.
How to set the rules for what an agent can do alone and what needs your sign-off, with every action waiting in one place before anything posts.
Why asking AI the same question twice can give you two very different answers, and the extra tweaks that turn a vague reply into an accurate one.
What it looks like when AI can see the contract, the vendor and the team behind every line, so you get the reason and not just the number.
How to turn a pile of vendor contracts into a prepaid schedule and the monthly entries in minutes.
Why the best recurring entry is the one nobody posts and pathways to automate recurring JEs.
Thursday, 22 October 2026
11 AM EST / 4 PM BST
2026-10-22T11:00:00-05:00