This as-told-to essay is based on a conversation with Georgian Tutuianu, a 36-year-old AI engineer at HubSpot, based in Boston. His identity and employment have been verified by Business Insider. The following has been edited for length and clarity.
Loading audio narration…
I started my career as a structural engineer. I went from traditional engineering, to software engineering, to AI engineering, which I’m doing now at HubSpot.
While I was a structural engineer, there was a lot of new technology coming out with machine learning. There seemed to be a lot of opportunities to integrate it in the field. So I went back to school at Northeastern University to get my Master’s in computer science.
This is part of an ongoing series about workers who transitioned into AI roles. Did you pivot to AI? We want to hear from you. Reach out to the reporter via email at [email protected] or secure-messaging platform Signal at aalt.19.
I started working as a software engineer in 2021 at a startup. Then I ended up joining HubSpot. I’ve always enjoyed using software and automation to make users’ lives better, so the job really resonated with me.
In my current job, I help marketers scale with AI. It’s cool because I get to work with people and see their faces when things go well and not so well. The dopamine hit of building the right thing is really fun.
Dedicating a résumé section to side projects
When it comes to AI engineering, you have to apply LLMs for whatever outcomes you want. Everything I’m learning about LLMs right now is on the job. It’s like the Wild West. Everybody is doing whatever, and the barriers to entry are as low as they’re going to be.
Being able to talk in the weeds on a technical level in my interview showed that I have definitely messed around with AI before. You can’t really fake that.
The way I showed that on my résumé was straightforward. I have a section for personal projects. I have a lot of side projects, so I took out what wasn’t relevant for this particular role. Then I added a project that I had been working on that was relevant.
Georgian Tutuianu
Something that I found super useful in day-to-day coding was having well-documented functions, which is even more important now because AI reads the entire code base. It turns out writing well-documented functions sucks because it’s very time-consuming, so I made the AI do that for me.
Now I just review them and rarely make edits. That was my personal project, which I still use. I only did that one AI project, but it was enough because it was a deeper and juicier project.
My side project came up naturally in the interview because I was asked to talk about a time that I used an AI agent or built an AI agent. It was very straightforward.
The interview wasn’t just project-based. I had to do a take-home coding assignment and meet with the hiring manager to review it afterward, and talk about the solution and tradeoffs. There was no algorithmic component to the interview, which is typical in software engineer interviews. I would characterize it primarily as focused on: “Can you build what we need?”
The best way to cut through the hype
I could have gone and found videos about the things I needed to learn, but the field is so diverse and fast-paced that it’s faster to try things out for yourself instead of waiting for video content or someone else to figure things out for you. There’s also a lot of AI hype out there so it’s difficult to figure out who is trustworthy.
The most useful part of my software career was learning systems thinking and seeing the big picture. If you understand what the system’s trying to do, you can take shortcuts and leverage the LLM in its best ways.
I think the best way to cut through all the AI hype is to try things out yourself. Go and investigate whatever you find interesting.
If you really love databases, go build some vector search tool and try to use it for a personal project.
It’s never been easier to just ask an LLM Claude or somebody to help you get started. It’s up to you to really build something fundamentally interesting that uses AI. That technical experience is valuable to people because you can immediately be useful when you start an AI role.
