Yesterday, a client asked about a detail from our meeting three months ago. Before AI, I would have spent an hour digging through notes. Instead, I asked Claude, and it pulled the exact quote, plus context from two related conversations I'd forgotten about. That's when I realized my AI assistant had become better at managing client relationships than my own memory. I’ll show you how I do it.
Claude Projects: Project Memory That Makes Sense
Every client gets their own Claude Project, where I store everything from proposals to meeting transcripts. When a client asks about a detail from three meetings ago, I don't have to dig through folders or rely on memory - I simply ask Claude. Before each meeting, I review the project status, and afterward, I upload the transcript and have Claude draft my follow-up email. Because it has access to the full project history, these summaries are thorough and contextual. This single change has cut my post-meeting admin time from 30 minutes to five.
And if you’re wondering how I get my hands on meeting transcripts: my AI meeting assistant read.ai joins all my meetings.
Want to try this? Start small: Create a Claude Project for your most important client. Upload your last three meeting summaries and ask Claude to identify patterns in their priorities and concerns. You might be surprised what it spots.
Custom GPTs: Your Specialized Task Force
While Claude Projects handles the complex stuff, custom GPTs are my go-to for routine tasks. Think of them as highly trained specialists who do one thing really well. My social media GPT knows my brand voice and hashtags, while my proposal GPT understands my pricing structure and templates. At $20/month for ChatGPT Plus, building these specialized assistants has been one of my smarter business investments.
Here's a helpful tutorial for creating your first custom GPT.
What Gets Measured Gets Managed
Speaking of smart investments - I finally started tracking my time in Harvest. The data changed how I think about automation. Those "quick" client check-ins? Five hours a week. Social media? Eight hours. Now when I build automation tools, I know exactly where to focus for the biggest impact. Sometimes the best productivity hack is simply understanding where your time goes.
In the next issue, I’ll finish up our series on Claude with a story on what Claude revealed about my writing style.
Until next time,
Jamie
There’s a reason 400,000 professionals read this daily.
Join The AI Report, trusted by 400,000+ professionals at Google, Microsoft, and OpenAI. Get daily insights, tools, and strategies to master practical AI skills that drive results.

