One solved a personal headache. The other tackled a marketing workflow. Both reminded me why curiosity might be the most important skill we have right now.
Election night. November 3, 2026.
I was sitting on my couch watching the results roll in, and apparently my brain decided that wasn't enough stimulation for one evening. So I built two AI agents. In 20 minutes.
One for my personal life. One for Dooap.
And before someone starts asking about my development background, technical certifications, or whether I know how to code, let me save you some time. I don't.
I know what I want to accomplish. I understand the processes that are making my life harder than they need to be. And I'm curious enough to see whether technology can do something about them.
That's it. That's the secret. And it's becoming a pretty powerful one.
Agent #1: My Calendar Finally Works for My Life
I travel. A lot.
For work, for fun, for speaking engagements, for visiting people I love, and occasionally because I've decided I need to be somewhere other than Ohio. And while I love being on the road, coordinating my schedule with everyone else has always been a pain.
I have a shared Google Calendar called "Sam's Travels." My family uses it. Friends who help take care of my dog, Ziggy, use it. My Columbus besties use it to figure out whether I'm actually going to be home long enough to make plans.
It's a pretty important little calendar. The problem? I had to maintain it manually.
Every time I booked a conference, committed to an event, extended a work trip, planned a vacation, or decided to sneak away for a long weekend, I had to remember to update another calendar. And let's be clear, my personal travel is just as important as my work travel. Whether I'm heading to Europe for a conference, taking a vacation somewhere fabulous, visiting friends, or booking a spontaneous getaway because I found a flight I couldn't resist, the people in my life need to know when I'm gone. And sometimes, more importantly, when I'm actually home.
Sometimes I remembered. Sometimes I remembered after somebody asked whether I could watch Churro, my neighbor's dog and Ziggy's boyfriend, grab dinner, or do something on a weekend when I was actually going to be in another country.
Not exactly a winning system.
So I opened ChatGPT and decided to change it.
I uploaded an Excel spreadsheet with my tentative Dooap travel plans for 2027. I asked it to identify the events with actual dates and put them on my shared Google Calendar. Then I started getting specific. Don't add events without dates. Make the titles consistent. Use the event name, a dash, and the location. Mark the days as busy. Add a travel day before and after each conference.
Done.
And then I thought, why am I still going to do this manually next time?
So I took it a step further. I configured an agent to watch for multi-day holds on my Dooap Microsoft Outlook calendar and reflect those trips on Sam's Travels. And because apparently I'm incapable of leaving a perfectly good workflow alone, I added another rule. If I'm going to Europe, build in an extra week so I can spend time with my Finnish team.
Because business travel is an opportunity to build relationships, not just collect hotel points.
And the beauty of having one shared calendar is that it doesn't have to stop at work travel. Vacations, personal trips, girls' weekends, and whatever adventure I decide to book next can live alongside my professional commitments. One place for the people who need to know where I am, without needing access to every detail of my work or personal calendars.
The best part? I tested the workflow, and it worked.
A task I'd been doing manually for years was suddenly something I could hand off. Not groundbreaking technology. Not some futuristic enterprise AI transformation. Just a real problem in my actual life that no longer needed all my attention.
That's the kind of AI adoption I want to see more of.
Agent #2: From Event Scans to Actual Nurture Campaigns
Now, let's talk marketing.
Because the second agent wasn't about getting my personal calendar organized. It was about eliminating the repetitive work sitting between a good marketing strategy and actual execution.
At Dooap, we've been working on our follow-up campaigns from Dynamic Communities events. If you've ever managed event marketing in the Microsoft ecosystem, you know the drill. You've got contact scans, different audiences, different conversations, different levels of interest, and somehow all those contacts need to turn into relevant follow-up instead of one giant email blast that basically says, "Thanks for stopping by our booth!"
No. Thank. You.
We can do better than that.
I already had the strategy. I had bullet points outlining the content, the stages, the audiences, and the messaging. I also had additional context from the data and my knowledge of the people we were trying to reach. What I didn't want to do was spend hours manually translating all that into five separate nurture campaigns inside HubSpot.
So I opened Claude.
Using the HubSpot connector, I gave it the information and asked it to organize the contacts and messaging into five distinct nurture campaigns. It built the outlines, and then I asked the next question.
Can you actually build them in HubSpot?
Turns out, yes.
Instead of stopping at a document telling me what I should do, I had Claude create the campaign structures in the system where the work would happen. Now, I'm going to QA everything before anything gets sent, because handing work to AI doesn't mean handing over accountability.
I'm still responsible for the strategy, the segmentation, the messaging, the customer experience, and whether those emails deserve to show up in somebody's inbox. But I don't need to personally do every repetitive step between the idea and execution.
And now I have a repeatable workflow I can feed future event companies and contacts into.
That's where things get interesting. Because the value isn't just the time saved on these five campaigns. It's the work I won't have to rebuild from scratch after the next event. And the event after that.
Stop Asking AI for Answers. Start Giving It Responsibilities.
I think we've spent far too much time talking about how AI can help us write faster. Write an email. Summarize a meeting. Draft a LinkedIn post. Give us ten ideas for a campaign. Fine. Those things are useful.
But that's not the most interesting part anymore.
The real opportunity is identifying the work we keep repeating and asking whether we should still be the ones doing it.
Think about how many times you've had the same process sitting on your to-do list. Export the data. Organize the spreadsheet. Segment the audience. Build the campaigns. Copy the details into another system. Update the calendar. Follow up with the team. Then do it all again next month.
Why? Because that's how we've always done it?
That's a terrible reason to keep doing something.
And I'm not suggesting we automate every human interaction out of our businesses. Quite the opposite. I want to automate the administrative nonsense so we have more time for the work that actually requires humans.
Strategy. Creativity. Judgment. Relationships. Trust.
The things that make people valuable in the first place.
You Don't Need to Know How to Code. You Need to Know How to Think.
Here's something I keep coming back to. I couldn't have built these workflows if I didn't understand the work they were replacing.
AI didn't magically know how I wanted my calendar managed. I had to explain my rules, my preferences, and the exceptions that matter to me. Claude didn't invent my event marketing strategy. I brought the audience knowledge, the messaging, the campaign stages, and the context needed to make those nurtures relevant.
That's the part I think people are missing.
The better you understand your business, your customers, and your own workflows, the more effectively you can use these tools. Domain expertise isn't becoming less valuable. It's becoming the thing that makes automation useful.
You still need to know what good looks like. You still need to know when the output is wrong. You still need to understand what should never be automated without human review. And you absolutely need to understand what data you're putting into these systems, who can access it, and what permissions you're giving an agent before you let it start touching customer records or company tools.
An agent that can execute a bad decision faster isn't progress. It's just a faster bad decision.
And Yes, We Need to Talk About the Environmental Impact.
I can't write about building agents for fun without acknowledging something else. AI has a real environmental footprint. The compute infrastructure, electricity, cooling, water consumption, and physical resources behind these tools matter.
And I'm not going to pretend that because I love experimenting with AI, those concerns somehow disappear. They don't.
I consume an enormous amount of information. I test tools. I ask questions. I build things. I challenge what the technology can do. But I also think we need to become more intentional about how we use it.
Not every task needs an elaborate agent. Not every question requires a massive model. Not every piece of content needs to be generated, regenerated, and regenerated again. There's a difference between useful experimentation and mindless consumption.
I'm still learning how to find that balance. And I think being excited about innovation and being responsible about its impact can exist in the same conversation.
They should.
Curiosity Is the Competitive Advantage Nobody Can Automate for You.
Twenty minutes. Two agents.
One that makes it easier for the people I love to know where I am. Another that helps me execute marketing campaigns at scale without sacrificing the strategy behind them.
Neither one required me to become a developer. Both required me to ask a pretty simple question.
What if I didn't have to keep doing this the same way?
That's where the magic happens. Not in the prompt. Not in the platform. Not in the latest shiny AI announcement. In the willingness to challenge the way things have always worked.
I see so many people waiting for their companies to tell them how to use AI. Waiting for training. Waiting for permission. Waiting until they understand every feature before they try anything. Meanwhile, the technology keeps moving.
You don't need to know everything. You need to know enough to start experimenting, enough to recognize when something isn't working, and enough to keep asking better questions.
I'm not interested in using AI to make more noise. I'm interested in using it to make more room.
More room for relationships. More room for ideas. More room for work that actually matters.
Curiosity runs in my veins. And I have absolutely no intention of turning it off.
So now I'm curious about you.
What are you building? What process are you tired of repeating? And, maybe more importantly, what's something you've been afraid to try because you convinced yourself you weren't technical enough?
Because I'd bet you're a lot closer to building it than you think.
