AI has completely changed what it means to be a software engineer. It also depends on how you become that person.
Before the AI boom, software engineering interviews were centered around creating code from memory, solving algorithmic puzzles, and demonstrating technical speed under pressure.
Now, CEOs are scouring GitHub and X for undiscovered talent. Increasingly, candidates are encouraged to use AI during interviews, emphasizing judgment and good taste.
For millions of software engineers around the world, that means navigating a fundamentally different job market than just a few years ago.
Le’ale Addison, a recent computer science graduate, experienced this transformation firsthand, interning at Amazon and KPMG before taking a technology internship at packaging company Smurfit Westrock.
AIs are so skilled at writing code that writing code has become commonplace in technical interviews. Karen Haberberg (BI)
Two years ago, the 22-year-old said she was asked to share her screen during a technical assessment to ensure she wasn’t cheating. Addison said that in fourth grade, students will be able to openly use AI and Google to search for answers in real time.
During her most recent job search, a few months ago, the employer peppered the interview with questions about AI. “How familiar was she with machine learning?” What about natural language processing? How was she using AI in her current workflow, and how would she use AI in her new job?
“These questions have never been asked before,” Addison said.
Addison’s experience reflects a broader readjustment unfolding across the tech industry, with companies like Dropbox and Cisco requiring engineers to demonstrate AI skills during the hiring process.
Anyone looking to break into an industry in flux or find their next opportunity faces a competitive market. According to HackerRank’s 2025 report, employment is on the rise, but 74% of developers are having trouble finding work.
Business Insider interviewed tech companies, AI startups, and career coaches to answer one important question: What does it take to land a coding job when AI is writing the majority of the code?
Goodbye resume, hello GitHub
Big tech companies like Google and Meta are competing for top AI engineering talent, offering dizzying pay and access to computing.
Meanwhile, startups are getting creative to attract potential hires. Executives at AI coding startups Cognition, Base44, and Replit say: Scout engineers regularly by searching X posts and GitHub pages. These two platforms allow engineers to post their latest projects for praise and recognition from their peers.
Stacey La Torre, the company’s chief human resources officer, said that at Replit, X is quickly becoming the “primary medium” for recruiting. Ordinary employees are also participating in the search.
Xavier Contreras, head of data engineering at a hedge fund, has been building, designing, and maintaining systems for nearly a decade. Karen Haberberg (BI)
“We have a Slack channel called ‘Talent Spot’ where people can basically say, ‘I connected with this person on LinkedIn, or I met this person at a conference or at X,'” La Torre said.
Emily Cohen, Cognition’s director of human resources and operations, said that all of her employees consider talent discovery to be part of their responsibility, and that she has personally traveled the world to persuade talented engineers to join the startup.
“Last week, I drove a candidate to the airport because I wanted to be the last person they talked to before they left San Francisco,” Cohen told Business Insider.
AI takes over technical interviews
That’s not the only reason companies are looking for talent. Vibe coding has changed the way employers conduct interviews and what they want to hear from candidates.
LeetCode is an online platform used to assess candidate skills and was once a rite of passage for candidates. Even if it is included, it is only the first hurdle.
Xavier Contreras, head of data engineering at a New York City-based hedge fund, has been building, designing, and maintaining systems for nearly a decade. He said his recent job search was much different than what he experienced five or six years ago.
At the time, he said, interviews were centered around coding challenges. During a recent job search, he was asked to explain his reasoning to an interviewer, including explaining the architecture behind a project he built and defending a technical decision. When given a take-home assignment, We are now allowed to use AI to do tasks that previously would have taken a month. For AI, the expected timeline was 3 days.
Contreras said AI has brought software engineering, data analytics and data science under the same umbrella. Karen Haberberg (BI)
Now that AI can do more of the heavy lifting, employers want to know that their employees understand how the technology works and can fix it if it goes off track.
Erin Scruggs, head of global talent acquisition at LinkedIn, says the most valuable technical skills today are systems thinking, problem solving, and the ability to quickly operationalize new agent systems.
“Companies need people who can interpret complexity, make decisions about how systems should work, and lead teams through rapid technological change,” Scruggs said.
This has led some companies to assign tests that provide a broader picture of a candidate’s skill set. Scott McGuckin, Cisco’s vice president of global talent acquisition, said the company is moving away from technical coding challenges to project-based exercises. The company also integrates AI-assisted development Bring the environment into interviews and observe how candidates function with AI-enabled workflows.
“The human element of oversight and expertise is more important than ever,” he said.
Searching for “data unicorn”
Transition to AI-assisted recruitment exam This does not mean that technical rigor is no longer important. Sundeep Teki, a career coach who helps staff staff state-of-the-art AI labs, said the technical hurdles are constantly evolving and getting higher when it comes to roles at coveted companies like OpenAI and Anthropic.
In addition to the technical know-how, AI knowledge, and soft skills needed to excel at work, companies are now placing greater emphasis on cultural fit. For example, he said, Anthropic includes specialized, non-technical cultural interviews for every role. He said candidates often aren’t hired if they don’t seem to identify with the company’s mission.
Is there another way to determine if someone will fit your team’s culture? Ask them to start the job.
Work trials, where candidates appear in person and collaborate with their prospective team, are becoming increasingly popular. AI coding startups like Lovable, Cursor, and Kilo all run trials as a way for both candidates and companies to see if it’s a good fit.
“We’ll be meeting with all of us in Amsterdam to get started in person, and we’ll be shipping a functional product the next day,” CEO Scott Breitenother said of Kilo’s latest international bootcamp. “Many people take their PTO out of their current job and come work with Kilo during our intensive week. Then, once they pass, they leave their job.”
One reason the hiring process looks different is because the job itself has changed. In the age of AI, many aspects of a developer’s traditional work are being automated and technical roles are merging, and employers are looking for candidates with broader and more versatile skill sets – jack-of-all-trades candidates who can adapt to multiple functions.
Contreras, a hedge fund data engineer, said software engineering, data analytics and data science have all moved under the same umbrella. That means you’ll need a different skill set to get and keep that job.
“They want to hire one person instead of three,” Contreras said.
As employers seem to be continually refining what they’re looking for, companies are looking for what Contreras calls “data unicorns.”
“I’ve been busy,” he said. “There’s a lot of information you need to know right now.”
