Stop Overthinking AI: What Recruiters Actually See When You Apply

TL;DR: Most ATSs struggle to parse basic resumes correctly, let alone act as sophisticated AI gatekeepers. Focus on clarity over clever workarounds. While everyone uses AI to write resumes now, the candidates who win are those doubling down on specificity and human connection.

After a decade recruiting for Meta, Google, Uber, and now health tech startup Rad AI, Emilio Arias has seen every resume trick in the book. White text stuffed with keywords. Graphics meant to bypass ATS filters. Resumes that mysteriously weigh 840 pounds (wait, that's something else entirely).

🎥 Watch the full session on YouTube

Here's what actually matters when AI touches your application.

The Reality Check: Your ATS Isn't That Smart

"Most of us are fighting our ATSs to just work the way we want them to," Arias told a packed Wellfound webinar audience. Forget the mythology about sophisticated AI gatekeepers. The truth is more mundane: recruiters spend more time wrestling with buggy software than deploying cutting-edge algorithms.

What these systems actually do: keyword matching and basic inference. The same thing ATSs have done for decades, just slightly faster.

What to do instead of gaming the system:
Write for clarity. Make your resume scannable for both humans and machines. Skip the text boxes, graphics, and columns unless you're applying for a design role. That invisible white text with keywords? Arias has seen it. It makes you look desperate, not clever.

Mirror What They Want (Without Being a Robot)

Read the job description. Actually read it and then mirror the language back.

If they want someone who "deployed ML models into production in the health tech space," your resume should say: "Deployed transformer-based NLP models to production for 5 million users, driving 30% growth in user engagement."

Notice the formula: Action + Specificity + Scope + Impact.

This isn't about keyword stuffing. It's about proving you understand what they need and have done exactly that. The difference between getting screened out and getting a call often comes down to using their exact terminology for the role.

The Fraud Problem Nobody Talks About

Here's something recruiters don't usually share: fraudulent candidates are everywhere right now, and it's making life harder for legitimate job seekers.

Arias revealed:

"We're seeing people show up to interviews who are clearly not the person from their LinkedIn profile"

The result? Automated fraud detection systems that flag normal behavior as suspicious.

Protect yourself from false positives:

• Keep your LinkedIn updated and matching your resume exactly
• Use a real photo on LinkedIn (yes, it matters)
• Turn off your VPN when applying, or set it to your actual location - VPNs can set off fraud filters in ATS systems and erroneously show you as a bad actor.
• Have some social footprint beyond just LinkedIn

That last point stings for privacy-conscious folks, but in this market, invisible candidates get treated as fake candidates.

The Human Skills That Matter More Now (Because of AI)

The irony of our AI-obsessed moment? Companies are doubling down on human qualities.

"I cannot tell you how many times a hiring manager spoke to a candidate and said, 'This person is maybe just a hair below what we need technically, but they clearly are hungry, they want to grow... let's hire this person.'"

Why? Because everyone has access to the same AI tools. ChatGPT can help anyone code. Claude can write anyone's emails. What separates candidates now is what AI can't replicate.

Critical thinking over credentials.

"AI can't touch your wisdom, your experience," Arias emphasizes. Can you synthesize information and solve problems, not just execute tasks?

Communication over coding perfection.

Those in-person interviews aren't going away. If anything, they matter more as deepfakes and AI-assisted interviews become common. "We're still doing in-person interviews, and I don't see that changing for the majority of the tech sphere anytime soon."

Adaptability over experience.

Companies need people who can evolve with rapidly changing tech, not just experts in yesterday's stack. "Continuously evolve. Show up to stuff like this. Lean on community more than ever."

What Makes Recruiters Actually Stop Scrolling

Arias's current search: machine learning research scientists. What catches his eye isn't fancy formatting or AI-optimized keyword density. It's specificity.

"Don't talk about what your team did. Talk about what you did," he emphasizes. Quantify everything. Not "improved website performance" but "reduced page load time by 40% through front-end optimizations."

Early career? Can't quantify impact yet? Focus on skills and learning velocity instead. Show what you're building, studying, shipping on the side. Prove you're actively growing, not just passively searching.

What AI Skills Actually Matter in 2025's Hiring Landscape

"How do we maintain a competitive edge in an AI-driven hiring world?" was the burning question from attendees. Arias's perspective, from recruiting ML scientists at an AI health startup, surprised everyone: Stop trying to compete with AI. Use it.

"You also have access to these tools. You can upskill and up-level yourself at a scale that you could not do before," he emphasizes. But here's what companies actually screen for: not whether you use ChatGPT, but whether you understand when AI helps versus when it hurts.

"AI literacy is important. Understanding what it can do for you, understanding how the models work, understanding how it can make your job more efficient." The candidates getting hired at AI companies aren't the ones with the most AI tools in their stack. They're the ones who can articulate exactly how AI amplifies their core expertise.

For Arias, who has ADHD and struggled with executive functioning early in his career, AI has been transformative:

"Where was ChatGPT when I was in college? The genesis, the start of projects, that's where I found the most effectiveness."

But he's quick to warn against outsourcing your thinking entirely. The winners will be those who use AI as a force multiplier, not a crutch.

Back to Basics: What Actually Gets You Past AI Screening

One candidate had everything: perfect experience, beautifully crafted resume, clear match for the role. They got rejected. The reason? They misspelled their email address.

Arias says:

"The basics still matter…Even in this age of AI, doubling down on them is just as important, if not more so."

While everyone obsesses over AI hacks and formatting tricks, here's what actually helps your resume survive automated screening:

Keep it scannable.

No text boxes, no columns, no graphics (unless it's a design role). AI parsing breaks with fancy formatting.

Match their language exactly.

If they say "Product Manager," don't list yourself as "Product Owner" even if the roles are identical. Algorithms search for exact matches.

Mirror your LinkedIn.

Discrepancies between your resume and LinkedIn trigger fraud detection systems. Keep them identical.

Include location basics.

City and state only. Your full address isn't needed and poses privacy risks.

Triple-check contact info.

That misspelled email means even if AI loves you, humans can't reach you.

His actual resume advice is refreshingly simple: Make it scannable. No columns, no text boxes, no fancy fonts. Use the exact job title from the posting. Keep your LinkedIn identical to your resume (fraud detection systems flag discrepancies). Triple-check your email address.

"You're overthinking it," he says about candidates trying to outsmart AI. The irony? Most ATSs can barely parse normal resumes correctly, let alone clever ones.

The "Will AI Take My Job?" Question Everyone's Afraid to Ask

When pressed about AI replacing workers, Arias offered historical perspective:

"Remember telephone switchboard operators? That was the number one employment choice for women in the United States. Automation killed it. We're still standing."

But he doesn't sugarcoat the current reality:

"There's something of a gold rush. Companies are trying to cut costs with AI. But once the hype subsides, companies will start to realize, okay, we actually need humans."

His advice for surviving the AI employment shake-up is counterintuitive: Don't fight the technology. Embrace it. "Someone's got to drive this ship. The robots are not there yet. AI literacy is important."

The jobs most at risk? Those focused on repetitive tasks without human judgment. The safest positions? Roles requiring what Arias calls "the human component": complex problem-solving, cross-functional collaboration, creative synthesis. "We're hiring you, the person. Not your ability to use ChatGPT."

The Portfolio Question

Should marketers include portfolios like designers? Should engineers share their GitHub?
*
"Don't be the one to eliminate information if you don't have to," Arias advises. "The filter's already going to do that for you."
*
If adding a portfolio or code samples takes five minutes and might move the needle 1%, do it. But only if it's relevant, well-organized, and actually shows your work. A half-built portfolio is worse than none at all.

Actionable Takeaways

• Create resume versions for each job title. A "Product Owner" applying for "Product Manager" roles should adjust their title to match what recruiters search for.
• Apply within 24 hours of posting. Set up alerts on Wellfound and other platforms. Early applicants make the shortlist that goes to hiring managers.
• Write your bullets using the A+S+S+I formula. Action + Specificity + Scope + Impact. Make every line prove you've done this exact work before.
• Keep showing up to webinars and communities. "That is 90% of it," Arias emphasizes. The people still actively learning and connecting are the ones who land roles.

See more “Ask a Recruiter" sessions on YouTube here.