Three editions in, and we’ve already seen AI prove its worth in finance functions of every shape and size, from matcha bars and dashboards to Power BI and SQL that writes itself (almost).
But this series isn’t just about cool tech. It’s about practical wins. It’s about solving real problems with the tools we already have and sometimes learning entirely new ones in the process.
So before we jump into the next wave of stories, let’s hit pause and reflect on what’s surfaced so far.
Common threads from the first half
Despite different industries, tools, and team structures, a few themes have come up again and again:
- AI won’t save you, but it will speed you up
Whether it’s Martin Goodwin teaching himself SQL and building an automated reporting system from scratch, or Luke Simpson automating Selfridges reconciliations at JENKI using OCR and Python, one thing is clear: AI doesn’t replace finance knowledge, it accelerates it.
Martin summed it up perfectly:
“It didn’t give me the right answer a lot of the time, but it gave me the start… and when it’s wrong, you still have to know, okay, why didn’t it work?”
For Luke, AI tools like Cursor helped him skip the Stack Overflow rabbit holes:
“I use Cursor to build things line by line. It helps me move fast, but I still do the thinking.”
The payoff? More time spent solving high-value problems, less time battling repetitive tasks or hunting for syntax errors.
- It’s not about replacing people, it’s about removing barriers
Phil Walker, co-founder of Mettryx, brings a refreshing realism to the AI hype cycle. He’s built a fractional CFO practice that uses AI to deliver enterprise-grade insights to SME clients, without the enterprise-grade budget.
But Phil’s cautious with the tools. He doesn’t chase shiny features. He uses AI when it removes friction, like writing DAX and SQL for Power BI, not when it risks undermining trust.
“You’ve got to have the domain knowledge to challenge what you’ve seen and whether it’s correct. It’s almost like coaching a part-qualified member of the team.”
And when it comes to AI writing commentary or client reporting?
“I’m yet to hear of anybody who’s really succeeding with AI on that. You can’t just give it a one-liner and expect magic.”
- Better inputs = better outcomes
AI might be new, but the old adage still applies: garbage in, garbage out.
That’s why Martin spent weeks building a solid foundation, connecting DataSights, cleaning the data, and learning exactly what to feed ChatGPT.
“The data has to be good before you can actually use AI. Garbage in, garbage out.”
When finance professionals take the time to clean up workflows and structure their thinking, AI tools become exponentially more powerful. But skip those steps, and you're just automating chaos.
- It's okay to start scrappy
One of the most encouraging threads across the stories so far? Nobody waited for a dedicated AI budget or perfect conditions.
- Martin had never used Power BI or SQL before. He just decided to learn.
- Luke built reconciliation automations using Python he’d taught himself years ago.
- Phil leaned into ChatGPT as a helpdesk, tutor, and junior assistant, giving it structure, context, and guidance.
In all three cases, AI wasn’t a moonshot. It was a practical step forward. As Luke put it:
“AI helps me work faster, but I still do the thinking. That’s the bit that matters.”
What’s coming next
The second half of the series is going to take things in a different direction, from behind-the-scenes automations to front-line impact, with use cases that stretch beyond the month-end close.
Here’s a sneak peek at what’s coming:
Mark Welton, CFO at Disaster Relief Australia
When disaster strikes, policies and procedures can’t afford to lag behind. Mark built a custom GPT to help him draft operational policies and onboard volunteers quickly. We’ll unpack how he scoped and trained it and how it’s already transforming documentation and training in a fast-moving, high-stakes environment.
Katy Johnson, Founder and CEO at AFI Balance
Katy supports multiple clients with a lean fractional team and she’s using AI to give them more than their headcount would suggest. From NotebookLM for research and internal knowledge management to Perplexity for digging deep into technical topics, Katy is redefining what a small but mighty finance team can deliver.
Final thought
So far, the biggest surprise of this series hasn’t been the power of the tools, it’s been the creativity of the people using them.
AI isn’t replacing finance teams. It’s amplifying them. Helping them work faster, learn quicker, and reclaim time for the work that really matters.
And we’re only halfway through.
If you're experimenting with AI in your own finance function or want to learn from others who are, jump into The Stack Exchange, our Slack community for finance leaders pushing boundaries.
See you next week.
.png?upscale=true&width=1196&upscale=true&name=Untitled%20design%20(42).png)
.png?upscale=true&width=1036&upscale=true&name=overview%20(4).png)
