By Loraine Lawson

Copyright thenewstack

During the hiring process, Skillsoft/Codecademy Enterprise Vice President Greg Fuller asks programmers whether they used AI to prepare a developer portfolio. This makes them nervous.

“As they’re hesitating, I’ll tell them my expectation is you are using AI to generate your projects, because if you’re not using AI, you’re not creating software with a modern mindset,” Fuller said.

Fuller and other hiring sources The New Stack interviewed said developer portfolios are still a good tool for landing a job, despite the growth of AI. In fact, they agreed that coding with AI should be shown as a skillset highlighted in portfolios — even for frontend developers.

How AI Changes Developer Portfolios

Fuller said AI changes the whole concept of portfolios and projects that developers create to enhance their resume. AI makes it possible to put together something “fairly compelling, fairly quickly,” he said.

Some hiring companies are now asking people to vibe code live as part of the interview process, he noted. While he hasn’t taken that step, he does think developers should showcase how they used vibe coding.

“Now, simple example is, let’s say I use a code base to create one project: You could showcase how you use AI to actually convert that to another code base, or two or three different other programming languages,” he said. “As you’re building these portfolios, be showcasing your adaptability, leveraging coding assistance, vibe coding or AI.”

Developers also should understand what technology was used and — perhaps more importantly — why they used AI.

“As you’re reviewing those portfolios, ask those probing questions: What GPT model did you use? Are you using a diverse set of tools, et cetera,” Fuller advised.

Portfolios become especially important for junior developers job hunting in today’s world, because entry level positions are being automated away, he added.

“Now I would actually disqualify people that say they don’t have anything to show anymore,” Fuller said.

A Frontend Perspective

Netlify’s senior manager of developer experience Sean C. Davis agreed that the role of portfolio websites is changing due to AI.

While portfolios still matter and can be a way to showcase creativity or highlight your skills in learning new technology, he thinks that AI code generating tools make portfolios a little less relevant, he said. While hiring managers once could look at a project’s code on portfolio or GitHub, now there needs to be a “paper trail.”

“The benefit that [portfolios] tend to bring folks probably looks quite a bit different today than it did five years ago, largely because of the influence of these AI code generation tools,” Davis said. “I tend to look at portfolios as not too dissimilar to how I might look at or talk about the value of a website or an application or anything like that, which is: is it achieving its goals?”

Davis, whose background is in frontend development, hires frontend engineers and technical developer relations workers for the company’s developer education team. The team is responsible for maintaining and building Netlify’s public web properties.

“The benefit that [portfolios] tend to bring folks probably looks quite a bit different today than it did five years ago, largely because of the influence of these AI code generation tools.”
– Sean C. Davis, Netlify’s senior manager of developer experience

Davis can tell if a website is AI-generated based on a basic one-paragraph prompt.

“The AI code generating tools today are contextualized by the developers that built them to default to certain styles, and so if a developer doesn’t bring the direction for UX/UI to that prompt, then you’re going to get that default sort of style,” he said. “If we go back, say 5-10 years, I’d attribute that to, they plugged in Bootstrap and then they didn’t make any customizations on top of it.”

Smart Portfolio Moves

Nowadays, Davis focuses less on code and more about whether the developer has the ability to build a refined and polished product.

“Where on the spectrum does this person fall, in the sense of to what extent did they use AI in the process? It could be near zero. It could be near 100 [percent],” he said. “What matters is that it at the end, we’ve got this polished piece, and that could come through a portfolio website, but it could also come through other side projects.”

The other piece he looks for is the developer’s ability to identify the problems they should be solving and how they go about solving those problems.

“We’re definitely looking for folks that are open to solving problems more productively than they have in the past, and today, that’s hinting at using AI tools,” he said. “I also care very much about the the art of frontend development, but also the level of refinement and polish that we’re bringing to these sites today, and I want to be really flexible about the way that we’re doing it. We’re finding what works for our team.”

Developer portfolios should be clear what the goals of the project are, he added.

Olga Lubiana works closely with both hiring managers and job-hunting developers as the technical teams business development manager at global tech talent firm Mobilunity. She’s seen first-hand what makes a portfolio stand out and what mistakes developers make, and she also sees how client expectations are changing due to AI.

It’s important for developers to be transparent about AI use, so that hiring managers can understand what was created by AI and what was custom-coded by the developer, says Lubiana. Developers need to be able to answer:

Why they decided to use AI in the project;
Where they used AI; and a
How they provided the information to the AI.

”The first and really important thing is to keep balance between using AI and using your own perspective, because hiring managers would like to hire humans and not machines,” Lubiana told The New Stack. “It’s about how you use it, how you approached it.”

Mobilunity also recommends developers use AI to create the presentation of the portfolio, with a caveat.

“It does not mean that it should be all copy-pasted from the AI,” she cautioned. “Instead, developers should keep a balance between their own thoughts and between some recommendation and style adjustments presented by AI.”