I've never been a financial advisor. That's why I built Meeting Notes Pro.

It started with one conversation. A friend who's a Canadian financial advisor had been trying ChatGPT because I'd been sharing what I was learning about AI tools. On a phone call, he described the problem: he could see AI's potential, but for the specific things he needed help with, it still felt very much like potential.

He'd been scrubbing client names and identifying information from meeting transcripts, uploading them to ChatGPT, but the output wasn't accurate enough. He had to check through the whole transcript and then edit the summary, and that took so long he may as well have typed the notes up himself. I took on that challenge because I like solving problems.

I built a CustomGPT and fine-tuned it until it actually worked for his process. It didn't work straightaway. It took a number of rounds of working with it, testing it, tweaking it again. But the final results impressed us both. It provided meeting notes in his preferred format and wrote client emails in his own voice and writing style. Where a big tech company might provide a tool that works, this one was being designed specifically for each advisor to work the way they want to work.

That tool now saves him eight hours every week. Hours he's using as planning time for his business. Time to work on his business instead of in it.

He actively told people about it. He showed them the process he was using, because he was so thrilled with it. Within weeks, all five advisors at another firm had adopted it. When compliance auditors saw the process and the results, they were impressed. What they specifically liked was the consistency of the meeting notes.

But I'd been carrying a niggle about compliance. The CustomGPT still ran on the ChatGPT platform, which didn't meet Canadian compliance requirements. And the conversations I was having with advisors revealed a more complex picture than simple security concerns. Some were waiting to see what their investment company would do. Some said their firm was already addressing AI, but from what I heard, those firms were improving back-office processes, not the client-facing work advisors actually needed help with. The pushback was strong. But the tool worked, and the need was real.

So I rebuilt it. That wasn't a quick or clean process. I first looked at Microsoft, but access to AI processed in Canada through Microsoft takes an enormous investment that only big companies can afford. I tried building a Copilot agent, but it couldn't be configured to the extent I needed. Then I found AWS, which has servers in Canada, including one in Montreal, which means Meeting Notes Pro can serve advisors in Quebec.

Meeting Notes Pro now runs on AWS infrastructure with a PII engine that removes all personally identifiable information on the advisor's own device before any data is transmitted. The AI processing uses Claude Opus. There were multiple rewrites, and this has not been a case of quickly vibe-coding a product. The code has been built with rigorous programmatic and user testing.

Through nearly a year of real-world testing with Canadian advisors, I discovered they faced challenges well beyond meeting notes. Compliance anxiety around AI tools. Vendors who didn't understand Canadian regulations. A widening gap between what advisors needed and what was available to them.

Northern Catalyst grew from those conversations. Not from a business plan. From listening.