I'm writing this from chilly Boston, about to head up to the woods of New Hampshire to spend the holidays with friends. Before I unplug for a few days, I wanted to take a minute to say thank you.

This newsletter has become something I didn't quite expect when I started: a real conversation. You've sent thoughtful replies, challenged my assumptions, shared your own implementations, and told me when something actually worked in your workflow. That feedback loop is what makes this worth doing.

As we head into 2026, I'm grateful for the time you've given me. I know your inbox is crowded and your calendar is worse. The fact that you open these emails and occasionally find something worth using means more than I can say.

2025 was the year AI in finance moved from "should we?" to "here's how we did it." We covered practical use cases, governance frameworks, vendor evaluations, and plenty of moments where the right answer was "don't use AI for this."

Some editions landed exactly right. Others I'd probably rewrite now. But every single one taught me something about what finance leaders actually need when they're implementing AI.

A few things stood out from our conversations this year:

The maturity level jumped. In 2024, we were still debating whether to use AI at all. In 2025, you were asking about audit trails for AI-generated analyses, building Custom GPTs for specific workflows, and designing control frameworks that actually work. That's real progress.

The vendor landscape clarified. We saw consolidation, clearer pricing models (mostly), and some genuine differentiation between AI-native solutions and legacy platforms adding AI features. The market is still messy, but it's getting easier to evaluate what actually delivers value.

The governance questions got specific. Not "do we need AI governance?" but "how do we validate AI output for regulatory filings?" and "what's the right approval workflow for client-facing AI content?" Those are the questions that matter when you're actually running finance operations.

The experimentation continued. Some of you built internal AI tools. Others tested vendor solutions for close processes or forecasting. A few let me share your workflows in the newsletter. That willingness to try things, measure results, and report back honestly — that's how we all get better at this.

A special thank you to those of you who took the time to reply to an edition, send me a note about what's working in your environment, or repost something that resonated. Those interactions matter more than you know. They tell me what's landing, what needs more depth, and what problems are worth solving together.

And to my paid subscribers: thank you for backing this work. Your support validates that there's real value in going deeper on implementation details, governance frameworks, and practical tools. It gives me permission to spend the extra hours building templates and testing workflows instead of just writing about them. I don't take that lightly.

I'm not making big promises about 2026 content. The best editions come from real problems you're trying to solve, and I don't know yet what those will be.

But here's what I'm thinking about:

More step-by-step playbooks for specific finance workflows. Less theory, more "here's exactly how to do it."

Deeper governance frameworks that let you move fast without breaking things.

Honest vendor comparisons. The market keeps evolving. I'll keep cutting through the noise.

And probably some surprises. AI moves fast. Your needs change. I'll keep showing up with whatever seems most useful.

Running this newsletter while working as a fractional CFO has been intense. There were weeks I wasn't sure I'd make the Tuesday deadline. But every time I thought about skipping an edition, someone would send a note saying a workflow saved them hours, or a template helped them explain something to their board.

So thank you for that. Thank you for reading, for engaging, for being patient when I'm learning out loud, and for making this feel like a conversation instead of a broadcast.

I hope you get some real rest over the holidays. The kind where you're surrounded by good people and not thinking about accruals or variance analysis. That's where I'll be — in the New Hampshire woods with friends, hopefully far enough from cell service that I can't check email even if I wanted to.

I'll be back in January with fresh ideas, more practical tools, and the same commitment to cutting through AI hype with real finance expertise.

Happy holidays. Happy New Year. See you in 2026.

— Anna