If there’s one thing my ADHD brain craves, it’s systems.
Beautiful, shiny, all-encompassing systems that promise to take the chaos swirling inside me and pin it neatly into rows, columns, and color codes.
So when I got the idea to build my own “executive assistant” out of bots, automations, and AI, I jumped in headfirst.
It felt like the perfect plan: no more lost tasks, no more missed approvals, no more “oops, I forgot.” Just me, my ideas, and a machine that never forgets.
At least, that was the dream.
The Dream System
I started with Telegram, because my team and I live there. Messages fly back and forth all day — approvals, updates, random ideas, urgent “ping-ping-ping” requests. The problem? By the end of the day, my brain couldn’t remember which of those were actual tasks and which were just chatter.
So I thought: what if I could just forward a message, tap a button, and have it automatically logged?
From there, things snowballed.
- Zapier would catch forwarded messages and drop them into Google Sheets.
- Each task would get a neat TaskID (T0001, T0002…).
- A Custom GPT would sit on top of everything, catching my brain dumps, tagging tasks with labels like payment, design, urgent, and pushing them into the sheet.
- Every morning, I’d get a task digest in Telegram. Every evening, a check-in reminding me what was pending.
- And when a task was done, I could just reply “done T0007” and it would magically update the sheet.
It was ambitious. It was elegant. And for a brief moment, it actually worked.
The First Wins
I’ll admit — those first tests gave me a dopamine high.
I forwarded a message from my designer asking for pamphlet approval. Zapier caught it, gave it a shiny new ID, and dropped it into my sheet as “Not Started.”
When I asked GPT to “log a task for Ancient Madurai: redo the pamphlet,” it neatly structured the entry and sent it off.
It felt like the chaos in my brain was finally being tamed. Tasks weren’t just slipping into the abyss. They had a place. A number. A home.
For about a week, I walked around smugly thinking: I’ve cracked it. I’ve built the ADHD system of my dreams.
The Cracks Appear
Of course, the honeymoon didn’t last.
First, Zapier started throwing errors. Something about webhook conflicts with Telegram. I spent hours digging into bot settings, privacy modes, webhook URLs — things that, let’s be honest, my ADHD brain has no business obsessing over at 11:30 p.m.
Then my beloved Google Sheet turned against me. My carefully written formulas kept breaking with messages like “Array result not expanded because it would overwrite data in A3.”
(If you’ve ever had a sheet scream at you like that, you’ll know the special mix of rage and despair it brings.)
Every fix spawned a new problem. I’d think I had it working, only to find tasks weren’t updating, or statuses weren’t mapping correctly, or GPT was sending empty payloads instead of actual task data.
Instead of reducing friction, the system became another giant task on my list:
“Fix the task system.”
When a System Becomes the Shiny Object
Here’s the thing about ADHD: we love building systems. It’s creative, it’s stimulating, it scratches the itch for control.
But systems have a dark side. They can become the project themselves.
Suddenly, I wasn’t moving tasks forward. I wasn’t checking off things that mattered. I was drowning in troubleshooting. My dopamine was tied to whether Zapier could parse a payload correctly — not whether I’d actually, you know, approved the pamphlet.
And the cruel irony? My shiny AI assistant was supposed to save me from that.
The Breaking Point
One morning, I opened Telegram expecting my neat little morning brief.
Instead, what I got was a broken message:
🌙 Evening Check-In 🌙
Pending Tasks:
Not Started: Error formatting not started tasks
I laughed. Then I wanted to cry.
Because that message summed up everything perfectly:
Even my system couldn’t tell me what was pending anymore.
Back to Basics
That evening, I did something I hadn’t done in years.
I put my phone down. I pulled out my old alarm clock.
No automations. No bots. No fancy dashboards. Just a simple, shrill reminder to wake up and look at a handwritten sticky note.
It felt humbling. After weeks of obsessing over AI-powered assistants and intricate workflows, I was back to the most basic tool there is: an alarm clock and a pen.
But here’s the thing: it worked.
What I Learned
Looking back, I don’t see the experiment as a failure.
It taught me three things that every ADHD entrepreneur (maybe every human) should remember:
- Tech is powerful, but it’s not magic.
No system, no matter how smart, can completely erase the chaos. And sometimes the cost of maintaining the system is higher than the cost of the chaos itself. - A system is only useful if it reduces friction.
The moment you’re spending more time fixing the system than using it, it’s time to step back. - Sometimes the simplest tools are the best.
My alarm clock doesn’t crash. My sticky notes don’t throw “array result” errors. And both of them keep me grounded in a way even the smartest AI can’t.
So, What Now?
I haven’t given up on systems. It’s in my nature to tinker, to build, to test. I still believe that AI + automation can be a game-changer for ADHD entrepreneurs.
But I’ve learned to hold my systems lightly.
To treat them as experiments, not saviors.
And to always, always keep a backup plan — even if that backup plan is a humble alarm clock on my bedside table.
Because at the end of the day, what matters isn’t the elegance of the system.
It’s whether I actually did the thing.