AI in financial advice: leveraging expertise and tech for better modelling

I had something of a lightbulb moment this week working with some very smart financial advisers applying AI to their businesses. 💡

We’re building AI features into their tech stack: extracting intelligence from unstructured advice documents, indexing content from the advice process for semantic search, deploying conversational analytics for firm-wide insights.

Really awesome capabilities that just weren’t possible a few years ago, and they’re genuinely valuable. They create opportunity.

But we’re also using AI to dramatically improve traditional processes in financial advice, replacing old tools that have needed replacing for years, but where the economics just didn’t make sense.

These features don’t so much create opportunity as remove pain, and that matters more.

One stark example of this is scenario modelling. Most advisers I talk to really don’t love the modelling tools they use. Often models can’t handle enough real-world complexity around entity structures, international jurisdictions, tax rules or “what if?” changes in market conditions.

Too many modelling tools don’t present well to clients and they’re clunky to use.

Advisers say “I know what I want here! I know the inputs, I know the maths, I know what the graphs and tables should look like. I could probably do it myself in Excel if I had the time.”

So we work together. They bring the domain expertise and the deep oversight of the calculations. I bring the tech skills and an army of AI agents.

Together we fast-track the requirements and design, I deploy proven frameworks and patterns to create a stylish and intuitive UI, then we build and test the functionality as a human-plus-AI team.

Sometimes I add an AI feature to automate data entry into the scenario model, which is a nice additional time-saver.

But simply fixing the long-standing pain point of a better modeller is what really moves the needle.

I like this way of working. It leverages the expertise we each bring to the table, and the result is super nimble and tailor-made to what advisers actually need. 🙂