Blog > Financial Planning > Train Your AI Assistant: A Practical Guide for Advisors

Train Your AI Assistant: A Practical Guide for Advisors

AI is already reshaping how advisors work.

It can save hours, generate drafts in seconds, and help you stay visible with less effort. No more blinking cursors. No more juggling every task yourself.

But there’s a reason many advisors hesitate. What if it gets something wrong? What if it sounds robotic or worse… non-compliant?

Those fears are valid and worth addressing.

Take the recent case of a New York lawyer who used ChatGPT to write a legal brief. The AI delivered six precedent cases in seconds. It looked convincing, but the citations had been entirely made up. The court fined the firm and threw out the case.

For advisors, the stakes are just as real. One unchecked message could confuse a client or raise red flags with compliance.

That’s why at Nitrogen, we’ve found it helpful to think of AI not as a computer, but as a junior assistant.

It’s smart, fast, and eager to help. But it doesn’t always understand your clients, and it doesn’t know when it’s wrong.

Like any new hire, it needs clear direction, regular oversight, and a final check before anything goes out. This is how AI becomes an efficiency tool, not a liability.

So how do you turn this mindset into action? Here are four techniques to make your AI assistant work for you clearly, safely, and in your voice.

1. Give your AI assistant real assignments

AI sounds helpful in theory, but one question can stop some advisors in their tracks: What exactly am I supposed to ask it for?

Like an intern, you don’t need to give it your most important job on day one. Start with small, useful tasks:

  • Draft a birthday note for a top client
  • Rewrite a blog post into a short email
  • Outline a LinkedIn post about recent market swings
  • Create a quick explanation of the Risk Number® for a new prospect

Then you can edit the drafts as needed.

This is one of the biggest benefits of AI. It gets you past the blank page.

Instead of starting from scratch, you’re reacting to a first draft. That saves time, reduces stress, and helps you get back into a creative flow more quickly.

Once you’re comfortable with the basics, Nitrogen’s AI Meeting Notetaker can take it even further.

It joins your client meetings (with permission), transcribes the conversation, and automatically generates a clean, simple summary.

It also creates CRM-ready notes and follow-up actions, reducing the need to toggle between apps or spend hours writing recaps.

You stay in control, editing and approving the output. But the busywork is already handled.

2. Give better direction, get better results

Most advisors treat AI like a search bar. They type in a question and expect a polished answer.

But AI isn’t built to find responses. It creates them from whatever you feed it. So if your input is vague, the output will be too.

Once again, think of it like managing a junior team member.

You wouldn’t say, “Handle this.” You’d say, “Write a short, reassuring email for a retired client. Keep the tone supportive and avoid industry jargon.”

The key to using AI well is learning how to give direction. Be clear, specific, and structured.

Here are some of the details you might want to provide in your prompts:

  • Who is it for? (e.g. a nervous retiree or a tech-savvy prospect)
  • What tone should it use? Calm? Upbeat? Friendly? Professional?
  • What format do you need? Email? Social post? Summary? Meeting Prep?
  • What context matters? Market movement? Client concerns? A recent meeting?

You’ll still need to review the result. But you’re saving time when not starting from scratch, and the quality improves dramatically.

3. Feed your intern the backstory

If AI sounds robotic, it’s not because the tool is broken. It’s because it doesn’t know you yet.

Most advisors give AI a basic prompt and get a basic response. But think about how you’d work with a new assistant.

If you just say, “Write a note about long-term investing,” you’ll get something that sounds ripped from a textbook.

But if you share a quick story, offer your usual phrasing, or explain the client’s situation, the results start to sound more like you.

Just like training an intern, the more background you give, the better the work gets.

For example, here are a few things you can feed into a prompt:

  • A short client story (without personal/confidential details)
  • An analogy you like to use
  • The tone you’re aiming for
  • A sample of how you’ve explained a concept before

Instead of asking AI to fill in the blanks, show it how you think. You’ll spend less time rewriting and more time polishing something that already sounds like your voice.

For example, try a prompt like this:

“Write a short email that explains the importance of staying invested during market volatility to investors with a Risk Number of 40-60. Use an analogy, like weathering a storm, to make the concept easier to understand.”

Clear context is what turns generic AI into a useful assistant.

Give it the raw materials, and let it do the heavy lifting.

4. Set boundaries your AI assistant can’t cross

Here’s the part that makes most advisors nervous: compliance.

You’ve seen what happens when AI gets it wrong. Maybe you’ve even tried creating something with ChatGPT and thought, “There’s no way I can send this.”

The fear is valid. In a regulated profession, the cost of getting it wrong is too high.

But here’s the good news: AI can be safe, but, once again, you have to treat it like any new assistant.

You would never give interns access to sensitive client information. And you don’t let them send emails without reviewing their work.

The same rules apply here. There are two simple guidelines to keep your AI use in bounds:

First, never enter client-specific or personal information into open models.

Most public AI tools don’t store data securely or privately. Once it’s in, you lose control.

A better approach is to describe a scenario in general terms. You’ll get what you need without risking exposure.

Second, always review before anything goes out.

AI doesn’t know your clients. It doesn’t understand your tone. And it won’t tell you when it’s making something up.

You’re still the editor. Before you send a message or publish a post, make sure it’s accurate, clear, and sounds like you.

You’re still the advisor. AI is just an intern.

Bottom line: AI can’t think for you or know your clients the way you do. But it can help you move faster, write more often, and spend more time on what matters most.

The trick, as we’ve said throughout this article, is to treat AI like your new office intern. It’s eager to help and surprisingly capable, but always needs supervision. With the right oversight, it becomes one of the most useful tools on your desk.

Looking for more help with getting started?

We’ve created an AI Prompting 101 guide with five ready-to-use prompts written specifically for financial advisors. Each one is built to save you time, reduce guesswork, and help you sound like yourself.

Whether you’re writing emails, social posts, or client updates, these prompts will help you put your AI intern to work quickly and confidently.

Download the guide and try your first prompt today.


Share This Story