When we started using AI in our practice, we made the classic mistake: treating it like a fancy search engine. Ask a question, get an answer, move on.
The real power of AI comes from building workflows—repeatable processes where AI handles the tedious parts while you focus on the work that actually requires a human advisor.
Here are the five workflows we use every week in our RIA practice. These aren't theoretical ideas; they're battle-tested processes that save us over 10 hours weekly. Combined.
| Workflow | Time Saved |
|---|---|
| Meeting Transcript Processing | 3-4 hours |
| Client Email Responses | 2-3 hours |
| Meeting Prep | 2 hours |
| Market Commentary | 1-2 hours |
| Content Creation | 2-3 hours |
| Total | 10-14 hours/week |
Meeting Transcript Processing
Every client meeting we have gets recorded and transcribed (we use Fathom, but Otter and others work too). The transcript itself is useful, but it's a wall of text that takes 15-20 minutes to review and summarize.
The old way: Read through 30-page transcript, manually extract action items, write up notes for CRM, create follow-up tasks. Time: 20-30 minutes per meeting.
The AI way: Paste transcript into Claude with a structured prompt. Get back: executive summary, client concerns/questions, decisions made, action items with owners, and CRM-ready notes. Time: 3-5 minutes per meeting.
With 8-10 client meetings per week, this single workflow saves us 3-4 hours. And the output is more thorough than what we'd create manually—AI catches things we'd skim over.
Pro tip: We built a Claude Project specifically for meeting processing. The custom instructions include our preferred note format, the types of action items to flag, and how to structure the summary. Every transcript gets the same high-quality treatment.
Client Email Responses
Client emails range from simple ("What time is our meeting?") to complex ("Can you explain why my portfolio is down when the market is up, and also walk me through the tax implications of that Roth conversion we discussed?").
The complex ones used to take 15-20 minutes to compose. You want to be thorough but not overwhelming, warm but professional, clear but not condescending.
The AI way: We've trained Claude to write in our voice (check out our guide on voice training). When a complex email comes in, we paste it into Claude with context about the client and key points we want to make. Claude drafts a response in 30 seconds that's 80-90% there. We review, personalize, and send.
Time per email: 3-5 minutes instead of 15-20. With 10-15 substantive client emails per week, the math adds up fast.
Important: We never send an AI draft without reading and editing it. The human review is non-negotiable. But the AI gets us past the blank-page problem and handles the structure so we can focus on accuracy and personalization.
Meeting Prep
Before every client meeting, you need to review: recent communications, portfolio changes, life events, previous meeting notes, and any action items that were supposed to be completed.
Gathering and synthesizing this information used to take 10-15 minutes per meeting—and sometimes we'd still walk in feeling unprepared.
The AI way: We pull the relevant data from our CRM and recent emails, paste it into Claude, and ask for a pre-meeting briefing. The output includes: relationship summary, recent activity, open items, suggested talking points, and potential concerns to address.
Time: 3-5 minutes. And we walk into every meeting genuinely prepared, not just "I skimmed the notes" prepared.
Bonus: For prospects, this workflow is even more valuable. We use AI to research the person (LinkedIn, company info, any public information) and create a dossier that helps us have a more relevant first conversation.
Market Commentary & Quarterly Letters
Clients expect you to have a perspective on markets. Whether it's a quick email during volatility or a formal quarterly letter, you need to communicate clearly about what's happening and what it means for their plan.
Writing these from scratch is painful. You stare at the screen trying to figure out how to explain something complicated in a way that's accessible but not dumbed-down.
The AI way: We give Claude the key market data and our perspective on what matters. Claude drafts the commentary, and we refine it to match our voice and ensure accuracy. What used to take an hour now takes 15-20 minutes.
For quarterly letters specifically, we've created a template in Claude that maintains consistent structure across quarters while adapting to current conditions.
Critical note: AI doesn't know what happened yesterday in the markets. Its knowledge has a cutoff date. We always verify current data and never trust AI for specific numbers or recent events. The AI helps with structure and explanation; we verify the facts.
Content Creation
Content marketing matters for advisors. Blog posts, newsletter articles, LinkedIn posts, educational materials—they build credibility and attract prospects. But who has time to write consistently?
The AI way: This very blog post started as an AI draft. We outlined the key points we wanted to make, gave Claude our voice profile and the target audience, and it produced a solid first draft. Then we rewrote, reorganized, and added our actual experiences and examples.
The AI handles the scaffolding; we add the substance. A blog post that would take 2-3 hours from scratch now takes 45-60 minutes.
The key: Don't publish AI output directly. It lacks your specific experiences, your client stories, your unique perspective. Use AI to get past the blank page, then make it yours. The result should be something only you could have written—AI just helped you write it faster.
Getting Started
You don't need to implement all five workflows at once. Pick the one that would save you the most pain this week and start there.
For most advisors, that's either meeting transcript processing (if you're drowning in notes) or client email responses (if your inbox controls your day).
Here's the progression we recommend:
- Week 1: Set up Claude and train it on your voice. Start using it for email drafts.
- Week 2: Add meeting transcript processing. Create a Project with your preferred format.
- Week 3: Build out your meeting prep workflow.
- Week 4: Experiment with content creation and market commentary.
By the end of month one, you'll have a solid AI workflow that's saving you real time every week.
Remember: The goal isn't to have AI do your job. It's to have AI handle the parts of your job that don't require you—so you can spend more time on the parts that do. More client conversations. More strategic thinking. Less time typing.
The Meta-Lesson
The advisors who will thrive in the AI era aren't the ones who ignore these tools, and they aren't the ones who hand everything over to AI without thought.
They're the ones who figure out the right balance: AI for efficiency, human judgment for everything that matters.
These five workflows are our current answer to that balance. Yours might look different. The important thing is to start experimenting, find what works, and keep refining.
The 10+ hours we save every week? That's time we spend on things AI can't do: building relationships, making judgment calls, and being present for our clients when they need us.
That's the real payoff.
Get our workflow prompts
Join our email list and we'll send you the actual prompts we use for each of these five workflows.