> For the complete documentation index, see [llms.txt](https://nexas-ridewiz.gitbook.io/lisaiceland/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://nexas-ridewiz.gitbook.io/lisaiceland/platform+/ai-knowledge+/ai-agents-in-finance-perspective.md).

# AI Agents in Finance (perspective)

* [x] Everyone’s talking about AI in finance. Here’s what I’m seeing on the ground.
* [x] Short term: there are no silver bullets in finance unless we fix the basics.
* [x] A chatbot summarizing emails or recording meetings isn’t transformation.
* [x] Long term: the hype is real. The upside is enormous.
* [x] Finance teams spend at least 40% of their time cleansing data.
* [x] We can’t get perfect data overnight. We start with what you have.
* [x] Focus on using agents to find anomalies and perform reconciliations.
* [x] Today’s business cases have poor short-term ROIs. 
* [x] Team up with your business partners. Our challenges are similar.
* [x] Compliance and Audit Committee approvals are hard. Find your best people to work on development before you move to production.
* [x] Don’t let perfection be the enemy of the good. Just start.
* [x] Give your teams access to the tools. Let them figure out where to focus.
* [x] Prompting is the foundation. Run “***promptathons***” to let your best people roll and your new learners, learn.