Everybody knows Cursor. Their AI-powered code editor lets anyone write and build software using only natural language. They've been on a meteoric tear, scaling from four founders in 2022 to more than a thousand employees by mid-2026 (not to mention an acquisition by SpaceX).

Their growth is the stuff of legend. But what you might not know is that their IT team has been just four people for a majority of the ride.

We recently sat down with Cursor's head of IT, Daniel Macias, and IT systems engineer Johnny Ireland, to talk about how a four-person team keeps up with one of the fastest-growing companies in the world. Here are the five takeaways worth stealing for your org.

Watch the full fireside chat here >

1. Futureproof your IT team by building the solutions, not just responding to tickets

Most IT teams are built to react. A ticket comes in, someone acts on it, then closes it. Repeat forever. Daniel Macias wanted the opposite.

"I didn't want to build a team of responders. I truly wanted a team of builders," he told us. "Every IT support issue is an opportunity to potentially automate."

That's why his first hire was a systems engineer. He set the expectation before anyone even started: "Don't come here thinking about just responding to a ticket. We really want you to come here and think about how you're going to engineer a solution."

Every request is an opportunity to spot a pattern and engineer a repeatable solution. Instead of constantly reacting to tickets, the IT team uses Console to identify patterns, improve systems, and prevent recurring issues.

After all, you answer a ticket once and you've helped one person. Automate it and you've helped everyone else.

2. Growth doesn’t have to mean more grunt work

Cursor blew past 750 employees on its way to more than a thousand, and the IT team stayed at four. That can only happen with tooling that helps the work scale.

Console now automates roughly 65–70% of Cursor's IT requests, with some weeks landing even higher. But it didn't start there, and it didn't improve until the team made the call to run everything through Console.

"Anything and everything we do needs to go through Console. Let's go all in."

After making that choice, their automated IT requests nearly tripled. The automation lets the team spend more of its time on infrastructure, employee experience, and the broader strategy required to support a rapidly scaling business.

In fact, Console now runs in about six other departments at Cursor, from finance to office operations to physical security.

3. AI models keep getting better, but context is what makes them truly useful

Better AI models matter, but Cursor’s team emphasized that it’s not processing power, but context, that makes AI truly useful. (Our co-founder Neal makes a similar argument in his recent op-ed)

When they weighed building a workflow inside Cursor versus Console, the deciding factor wasn't raw model quality. It was that Console already knew their world. "Console just had more of the context of our systems already than Cursor did," Johnny said. Recreating that, the systems, the permissions, the workflows, the identity structure, somewhere else would have cost real time.

As Danny said, "I really do believe that the truth lies in the context."

Models will keep getting better. But an agent that understands your tools, data, and how your company actually runs will beat one without context, every time.

4. Use natural language tools to quickly prototype and deploy new IT workflows

Cursor's team uses natural language to quickly prototype, deploy, and optimize new IT workflows.

For one network troubleshooting playbook, Johnny built the first version fast. He described the objective to Console's assistant in natural language, told it what the workflow should look like, and asked what it needed from their network monitoring system. He had a working version in about an hour.

"I didn't sleep that night," he said. "I just wanted to keep going and solve the problem and expand it to see where else it could go." The team then improved it using real employee requests and feedback.

This creates a new operating model for IT: launch an initial solution, observe how it performs, and continuously refine it. Workflows no longer need to begin as large engineering projects. They can start as rapid experiments and evolve over time.

5. AI won’t eliminate the role of IT. It will make space to work on more valuable problems

"Have no fear. AI is not going to take your job. We have plenty of work," said Danny. What AI is doing is moving the work up the stack. Cursor retired the old "support rep" title entirely and rebranded the role as IT operations engineer, a blend of employee support and genuine tier-three and tier-four systems work.

Once the repetitive tier-one and tier-two tickets are automated, the humans get to spend their time on infrastructure, new systems, and driving AI adoption across the company. (There's still plenty of scut work for humans, too. Nobody's automating a faulty monitor or bad cable.)

"This is probably the easiest time ever in technology history to learn whatever you want," Danny said. Tools are everywhere, you can sign up for free developer accounts, deep dive into technical niches with a twenty-dollar AI subscription. And there's a huge network of people that are more than happy to help.

In other words, the ceiling has never been higher for someone willing to build.