eSkill

Barry the Capybara had it all figured out. Barely briefed on the basics of being a budget analyst, but burning with the belief he belonged in the role, Barry had a plan. He was going to take his hiring test in one tab on his laptop, and look up answers (and maybe ask ChatGPT for some help) in another. A bit of copy, a bit of paste, and boom — Barry believed he’d breeze by. Barry boots up the test… and bam! His browser is bolted into full-screen mode. On top of that, he can’t CTRL+C, CTRL+V or right-click on any of the test material. All of it is blocked. He can’t even bang out a CTRL+P to bring a printed backup to his buddies! And just like that, Barry’s plan is busted. Beat back bad behavior before it begins. Block copy-paste. Bind browsers to full-screen. Ban right-clicking. Build a balanced, bias-free, and bulletproof testing baseline for every budding candidate. Browse additional features of anti-cheat here: https://lnkd.in/gHsAFveT

Last year, Anthropic CEO Dario Amodei said AI technology could disrupt half of entry-level white-collar work — and recently, Anthropic mapped out the jobs that AI could possibly replace: among them financial analysts, software developers, customer service reps, and computer programs. The least exposed jobs are those that are primarily blue-collar and consist of hands-on work, like warehouse worker positions, welding, and electrical work. Soon enough, the only jobs that’ll be available are one that truly only humans can do. If you develop these hands-on skills, not only will you ensure the likelihood of long-term employment, but you’ll probably also be prepared in the event an apocalypse sets civilization back several hundred years. Which is why we’re introducing our Medieval Test of the Month series. This month’s test: Swordsmithing! If the future is uncertain, you might as well be prepared to forge your own weaponry (or hire someone who can forge your weapons for you)! In this test, we cover blade metallurgy, forge safety, heat treatment and tempering, rust prevention and maintenance, and sharpening techniques. Get ready to duel your way through this next century. Next month, we’ll cover our butter churning test! (just kidding — Happy April Fools!) Check out the swordsmithing test (or hundreds of tests for real, in-demand roles) here: https://lnkd.in/gNU5jSEu

Let me share a use case where I think a former customer went too far with AI.   We worked with a customer that seemed to be chasing the shiny object. The assumption was: if AI can generate content quickly, it can generate assessment content too. And so, after many years of fruitful collaboration, they decided to go with a new platform that had all the right buzzwords: conversational AI recruiting, AI testing, AI grading, AI conversation bots.   AI can be incredibly useful for hiring teams. We’re big believers in that. But when you hand AI full control over something that requires relevance and factual accuracy — especially in hiring — you’re taking a real risk.   Here’s the thing. Assessments need to be validated. That’s why we collaborate with SMEs to produce our test content. Skillsets for different fields evolve in real time, and not everything is on the open internet for an LLM to scrape. If no human is reviewing the content, challenging it, and confirming it’s actually correct, you can end up building an evaluation process on a false foundation — or just creating assessments that are too surface-level for the actual demands of a position or the field at hand.   To reiterate: If you’re just scraping the loudest or most frequent answers off the internet, you CAN’T create good test content, at least, not with confidence. Good test content comes from subject matter experts with years of experience who understand the role, the field, and what competence actually looks like.   But because AI is often so confident in the way it presents information, it’s easy for people to trust outputs that shouldn’t be trusted. Automation bias is a real thing. Humans tend to trust computer-generated information over their personal judgment, even when the generated information turns out to be flat-out wrong, contradictory, or heavily biased.   Yes, AI can support the process. But no, it should not be the final authority on content that needs to be factual and defensible.   AI is a powerful tool. But in hiring, substance matters more than speed. If you’re not validating what AI produces, you’re not improving the process. You may just be automating bad decisions. There are times for full AI automation, but there are times for human in the loop.

One-way interviews — where candidates record responses to on-screen or AI-delivered questions on their own time, without a live interviewer — have their proponents and their critics. Some companies, like Spark Hire, market their platform as asynchronous one-way interviewing where AI is optional, while companies like HireVue use AI for analysis and feedback. Yet others, like Alex, are run entirely by AI avatars. None is without their controversy, although some platforms invite more risk than others. Read the full blog here: https://lnkd.in/gUnsH34R (tweet cred goes to Adam Karpiak on X)

Every hiring decision in government carries weight. A single hire can impact service delivery, compliance, team morale, and the communities agencies serve. And when a role needs to be reopened, the ripple effects are felt immediately. The problem is that most early-stage screening still leans on resumes. Resumes show experience. They rarely show performance. That's why more agencies are beginning to shift from asking, “Where has this candidate worked?” to asking, “Can this candidate do the job?” And where skills-based hiring makes a difference. We partnered with eSkill to explore how objective, job-related assessments are helping public sector teams get there: ⚖️ Reduce bias in screening 📈 Improve overall quality-of-hire 🎯 Identify qualified candidates earlier ���️ Support defensible hiring decisions Learn more in guide: https://lnkd.in/g84-a6Yy

We are so excited to work with Hebba Youssef and Workweek to help spread the word on why AI CAN be used in hiring…but should be approached cautiously and ethically! Give the article a read! (linked in post comments)

Our AI-Powered Proctoring feature, in partnership with Proctorio, provides powerful, professional, university-grade proctoring for remote candidate testing. Like for Polly the Capybara. Polly was applying to a payroll specialist position, but didn’t know anything about popular platforms in the field! No QuickBooks, no Paychex, nothing. Still, Polly presumed she had it perfectly planned. Her pal Pesto, a practiced payroll professional, would pose as Polly and perform the test himself (since the two of them are practically identical capybaras.) But that ploy fell through when they realized the test was equipped with eSkill’s ID Verification — so Pesto couldn’t pretend to be Polly, despite a perfect planned performance. But they soon developed another plan. Pesto could provide prompts from the periphery, and Polly could put his answers into place on the testing platform. But AI-Powered Proctoring caught the video of Polly looking off to the side plus the voice of Pesto telling her the answers! And just like that, Polly’s performance was properly penalized. Sorry, Polly — no payroll position for you. It’s back to accounting class you go! Peek at more features of AI-Powered Proctoring here: https://lnkd.in/gPrkFs3M

“We do an employee appreciation event every year,” Rachel says. “And while at that event, one of our long-term employees came up to me and she said, ‘What are you and Lisa doing differently?’” she recounts. “‘What are you talking about?’ They’re like, ‘We feel it. The people that you’re hiring are so much better.’” We had the absolute privilege of speaking with talent specialists Lisa Nicholas and Rachel Gustofson at nonprofit CCRI (Creative Care for Reaching Independence), who shared with us how eSkill helps them hire better employees to provide a better quality of life for individuals with disabilities in Minnesota. Read the full case study here: https://lnkd.in/gnXPpEzT

Clarence the Capybara was clearly about to cheat on his hiring test, conceding that his capabilities weren’t quite cut out for the client coordinator role he confidently cast his resume for. Cut off from copy-paste and right-click (courtesy of anti-cheat controls enabled on the test), Clarence clutched his phone and cast a cautious, covert glance down into his hands — calling on Claude to conjure a clever response. Wait — what’s this? Clarence’s sneaky scheme was swiftly spotted by eSkill’s AI-powered proctoring solution! Caught red-handed, Clarence was cast out of the candidate pool — his sky-high suspicion score, along with screen and session footage, carefully considered by the recruiter. And it’s all thanks to eSkill’s suite of proctoring tools. With our latest updates to proctoring and anti-cheat, it’s time we provided you (and resourceful test takers) with a refresh. eSkill’s proctoring solutions — Anti-Cheat and AI-Powered Proctoring — create a secure testing environment that levels the playing field for test takers, so candidates like Clarence can’t cheat their way into a job. Among people taking creative liberties on their resumes or calling on AI to ace an interview, assessment security safeguards like proctoring save you from the steep cost of hiring — and then having to fire — the wrong fit. A catastrophic call, Clarence. Curious? Check out more about proctoring here: https://lnkd.in/gBdzkNhN