The other day, I was trying to build a demo site in Drupal using an AI agent. My instruction was simple: “Do what you need.”
Figure it out. Make the decisions. Get it done.
It didn’t work.
The instruction was too broad. The agent didn’t know where to start, what to prioritize, or how to sequence the work. Once I broke the task into smaller pieces and tackled them one at a time, things started falling into place.
And somewhere in that process, a memory surfaced.
I’d been through this exact same thing before — except the agent was me.
In 2019, I joined Acquia Japan as a founding member. At the time, only two companies in Japan were using Acquia’s products. Today, that number is over fifty.
On one of my first days, my manager gave me a single instruction:
“Do what you need.”
That was it. No playbook, no detailed task list. Just a mandate to figure it out, report back on what I’d done, and ask for help when I needed it.
Honestly? I had no idea what I was doing.
I was an engineer. I’d been given the title of Presales Engineer, but I didn’t know the first thing about sales. I couldn’t build a quote. I didn’t know how to run a customer meeting. I didn’t even know what I didn’t know.
What saved me was my teammate — the Account Executive who joined as a fellow founding member. To me, this person was something closer to a mentor.
What does it mean to earn a customer’s trust? How do you explain complex technology in someone else’s language? How do you listen before you pitch? I learned all of it by watching them work, side by side, day after day. If I’ve grown from a solo engineer into a Lead Solutions Engineer, it’s largely because of them.
There were countless late nights spent messaging headquarters across time zones. The jet lag of asynchronous communication was real. But speed mattered more than sleep. If I didn’t know something, I researched it. If I couldn’t find the answer, I asked. And whatever I learned, I shared immediately. The global Presales team was incredibly supportive — and being able to meet them in person at the Boston office right after joining made a real difference.
Localizing materials wasn’t just translation. It meant rethinking entire presentations so a Japanese audience could grasp the full picture on a single page. Demos couldn’t just be replayed from the global playbook — the storytelling style didn’t land the same way. I had to rebuild them around local use cases to overcome the perception that this was “just a foreign product.” If a resource didn’t exist, I created it from scratch. A quick diagram here, a detailed one-pager there. Those small, unglamorous efforts added up to something bigger: trust.
Looking back, the things that worked all had something in common.
Nobody limited themselves to their job description. I staffed the registration desk at events. Not once did anyone on the team say, “That’s not my job.” We joked around, we leaned on each other, and we moved forward together. That culture — scrappy, generous, all-in — defined our small team.
Every manager I’ve had, the colleagues who joined in those early days, our GM at the time — they were all people who showed up, pitched in, and had each other’s backs.
I didn’t accomplish anything on my own. I was simply surrounded by the right people. And for that, I’m grateful.
Back to the AI agent that couldn’t handle “Do what you need.”
Breaking a big, ambiguous mandate into small, concrete tasks and tackling them one by one — that’s the lesson the AI taught me. But it’s also exactly what I did six years ago, long before I’d ever heard the term “agentic AI.”
If I had to do that startup phase all over again today, I’d go all in on AI. The speed would be incomparable to what we had back then.
But speed isn’t what carried me through those early days.
It was a manager who said “Do what you need” and meant it. A teammate who showed me what trust looks like in action. A global team that answered my midnight questions without complaint.
AI is a powerful tool. But the people who believe in you? No tool can replace that.
To everyone who was part of those early days — thank you.