Hey Fintech Nerds 👋

I’ve spent the last couple of months obsessed with how companies are using AI to transform their businesses. Now I’ve put all of those insights, conversations, and threads into a single report that covers how fintech companies, banks, and even AI labs are using AI to do more.

👇 You can download the report here for free, or view it natively on our website here!

Here’s a quick taster of the 50+ page report.

Every CEO and strategy team I speak to wants to know how other people are using AI. Because the dirty little secret of AI is that, despite it being the fastest-growing enterprise product and rollout in history, the vast majority of companies have seen no discernible benefit. Worse, some are simply getting an exploding AI bill, without the outcomes to show for it.

Most companies are stuck at level 0 or level 1. If users have an AI chatbot, but they can’t collaborate with colleagues, and there’s no system around them, there’s no way they get the productivity benefits. Instead, they get a blank chatbot box, and a lot of uncertainty.

The report profiles Anthropic, Ramp, Allica Bank, and BBVA (with a brief summary of JPMorgan and RBC) across a mix of Fintech companies, regulated banks, and incumbents. And it creates several principles

Anthropic collapsed their product release cadence from months to weeks to days, and rebuilt their entire operating model around it. 

Ramp has been ultra public on X about how they’re changing everything they do around AI, which made my life writing case studies easier. Why would they give so much away? Two reasons. 1) The fastest sprinter never looks over their shoulder; they just run fast, and 2) Hiring. It’s all designed to get talent to want to work there.

Allica is a UK bank most of you may not have heard of, but being a fully regulated bank and the second fastest-growing company in Europe for 2024 and 2025, they’re a more than worthy case study. Especially if the internal objection is “sure, non-banks can do it, but what about regulated banks.”

I needed a big bank, and BBVA has been the most public of the big banks about what they’re up to.

They’ve leaned into Enterprise OpenAI, which gives data privacy assurances (to a GDPR regulated firm), and plugs into existing back-end data sources. Users can graduate to being able to use AI with internal training, and then share custom “GPTs” ChatGPTs precursor to skills. But the team shape hasn’t materially changed but every squad now has an AI capability baked in. While some teams' operating models have, this isn’t universal across the company yet. 

Both RBC and JP Morgan have built sophisticated AI labs and are training models for internal use cases. Their implementation also fully manages cyber-risk and gives employees the latest data. But they’re as yet not able to share skills in the way Ramp can, and compound the productivity advantage. 

The hardest thing to change is your operating model. Smaller companies have built custom harnesses, skill sharing, and an entire employee curriculum and progression. The entire company aligns to this new operating model. Larger companies are like large shipping vessels that take longer to turn. Yet each of them competes with the younger companies on some dimension. 

With the help of AI and my burgeoning team at Fintech Brainfood, we’re putting together an all-star team to help companies actually get productivity out of AI. We’re staffing for solutions eng and FDEs and happy to chat if you hit reply.

If you want help telling your story, Fintech Brainfood and the Tokenized Podcast are open to high-quality sponsors and partners. We’re most interested in long-term partnerships rather than small ad buys, to support our world-class content (hit reply).

And, if you’re into this stuff, you should absolutely come to this year's Fintech Nerdcon in San Diego on the 19th and 20th November. We’ll have Seb, who built Ramp’s harness, doing a 60-minute masterclass on how they built it. We’ll also have Dennis from Chime doing a “How I AI” session for people who work in finance (in easy-mode and operator-mode versions). This will be the show where you learn how the rest of the industry is using AI, and who the best companies are to work with.

So grab it here while you can for free, before it goes into the archive

Was fun to make, and I owe a lot to Pet (of Brainfood / Tokenized producer fame), who essentially took my writings, my feedback, my ideas, and made this deck…with no design resources. Brainfood’s output is a living example of how AI can help us do more.

I hope you got something from this report.

And hit reply if you’d like to do more, and I welcome any and all feedback!

ST.

(Except: “Hey, why didn’t you write about my company” - don’t be that guy).