One of the most common mistakes I see in the job-seekers I coach is that they are trying to be everything to everyone.

You know, when someone’s LinkedIn headline says Data Scientist | Machine Learning Engineer | LLMs | AI Engineer.

I get the logic; You can do all those things, and you don’t want to rule yourself out of a job. But it’s probably not having the effect you want.

I recently asked a tech recruiter about this in an interview for my book on ML careers. She’s been recruiting for tech companies for about 20 years, and explained that trying to show up in every search can actually work against you:

The goal shouldn’t be to show up in as many searches as possible as a mid candidate.

Instead, we want to show up in the right searches as a strong candidate.

This isn’t unique to the job search. In business, we choose a specific customer and build something that serves their specific needs. If I’m opening a bakery, adding sushi to the menu probably won’t convince more people to come try my croissants.

Or even dating. If I say I love everything and am looking for literally anyone, I’ll most likely be forever alone.

The job search is more similar to sales than many of us would like, so never underestimate the importance of branding.

Rather than keyword-stuffing, choose a clear direction and make it easy to see how your experience supports it.

I find the idea of being a “T-shaped engineer” useful here. You go deep in one area and build enough breadth that you’re competent in everything else that’s relevant for your job.

For an AI engineer, that depth might be AI evals or building agentic systems for a particular industry. You still need software engineering, some ML fundamentals, and to be relatively data-savvy, but those are supplements to your core competencies.

Then make that specialization super clear across your profile. Use your headline to name the role you want and the work you focus on. Something like “Software Engineer building agentic systems” makes it really easy for me to see what kind of role you’d be a good fit for. Show the same narrative in your resume and when talking to recruiters.

You can change direction later. For now, make it easier for someone to understand where you’d fit and what you could help them with.

Building an AI agent on your laptop takes 10 minutes. Getting it to reliably handle real customers in production is a completely different can of worms.

In this week’s video, I walk through a demo of what a production-ready agent looks like using Google Cloud Platform. We’ll go through adding long-term memory, agent environments, permissions, evals, tracing, and more!

I also tried a new editing style on this one. I hope you like it.

Or read the companion guide here.

I'll be speaking at AI Dev Craft in Las Vegas, October 27-28th.

I'll be doing a session on Staying Current with AI Without Losing Your Mind during the main event, as well as a full-day workshop on Evaluating AI Systems on October 26th.

Learn more about the event here: https://ai-devcraft.com/

And my audience gets $200 off with the code: specialguest

Hope to see you there!

Forwarded this email? Sign up here.

Follow me on YouTube | Instagram | X | LinkedIn | TikTok

Note: This email may contain affiliate links. If you make a purchase I may make a small commission, at no cost to you. Thank you for your support!