RSS Amplifier

Blog

AI In Finance

Helping CFOs turn numbers into stories and build influence.

finstoryai.substack.comSource feed ↗10 posts

Live Last read · last published · next check

Written by

Latest posts

OpenAI's CFO Sarah Friar just published the AI finance playbook

OpenAI's CFO still closes the books manually. But she wants a zero-day close and continuously updated forecasting. 5 lessons for CFOs building an AI native finance function.

How to do financial modeling with Claude

Claude is the new Excel. Finance people can use Claude to do advanced financial modeling and build a storytelling board deck.

Why AI fails in finance

MIT warns CFOs. 95% of AI is dangerous.

Anthropic just released Claude Opus 5 and it's insanely better than Fable 5

Anthropic released Opus 5 yesterday. For finance this one is different. Fable 5 intelligence at half the price. Higher accuracy on financial modeling. And 841 connectors to your finance stack.

Claude just changed FP&A forever

We spoke to a CFO who asked us if they could build their entire annual budget inside Claude.

Fable 5 vs Opus 4.8 in finance: Only one survives the Audit

Testing Claude Fable 5 vs. Opus 4.8 in finance. We ran the hardest finance test we could design. The models disagreed. The lesson for CFOs is when to use which model.

Using Claude Fable 5 as a CFO to rescue a broken month-end close

Can Fable rescue a broken month-end close? And why it's better than Opus 4.8 for complex finance tasks. This is the finance-grade AI workflow nobody is talking about.

They are lying about AI in finance

Are we ever going to get ROI from AI? The AI feature you bought is a 90-day melting ice cube. MIT says 95% of AI produces zero measurable value. Anthropic's own product launches are erasing the...

Urgent Call for CFOs who want AI

Your ERP records financial reality. It does not explain it. Every month, finance rebuilds the financial story outside the system. That changes now.

How to build a real Financial Story with Claude MCP

We combined an MCP server with a Claude skill so the model pulls real data from NetSuite, analyzes it, and asks you 3 sharp questions per page before writing a word. Your judgment goes in.