Using artificial intelligence to streamline the hiring process is a wonderfully smart move for business leaders. But what's the use of paring down candidates only to find that the candidates the A.I. picked still don't have the skills you're after?

Enter Vervoe. Unlike other A.I.-based hiring systems, Vervoe focuses on using A.I. to automate the grading of skills-based assessments. Candidates get a chance to show off their talents and knowledge through real-time tests that mimic what they'd really be doing for you. The system helps you evaluate their performance and predict how well that result would translate to the candidate's ability to meet your needs.

Addressing what most A.I. hasn't fixed

Ben Eubanks, Principal Analyst at Lighthouse Research & Advisory, notes that employers already are exploring recruiting options like chatbots and automated sourcing tools. But he sees a glaring flaw in their approach.

"The challenge [...] is that this looks at hiring as a pure volume problem. Vervoe takes a different approach, focusing on the actual screening and decision-making process that is inherently flawed by human bias and making it more insightful."

Eubanks says that, in a 2017 study, his firm found that both candidates and employers agree that work simulations and skills tests--the heart of Vervoe's system--are the best predictor of how somebody actually will do on the job. And he says that, while candidates despise tests that seem disconnected from the tasks they'll really have to handle, they actually want to participate in skills assessments to display what they're capable of.

Accessible to everybody, ideal for SMBs

Vervoe's CEO and Co-Founder, Omer Molad, says that employers can test candidates for any skill, and that the A.I. can analyze any test. That makes the system relevant for virtually any company. But employers who are still growing might find Vervoe especially helpful.

"Hiring managers in smaller companies are often faced with the task of filling roles they are not necessarily experts in," he explains. "High-quality hiring technology may also be out of reach for SMBs. We're changing that by democratizing candidate evaluation and making A.I.-powered skill testing with expert content available to small and growing businesses."

And the testing process is fast. Molad says most customers start hiring immediately and have candidates who have completed skills tests (Talent Trials) the same day or within 24 hours. The tests typically don't take candidates more than 20-40 minutes at the first stage, although more advanced testing later on in the hiring process can take longer.

Molad's team claims an accuracy rate of 83 percent (this can get better every time you manually grade an assessment to teach the A.I.), and every candidate gets a chance to interview. And so far, 4,000 companies--including global brands like Allegis, Naspers and AKQA--in over 70 countries have put Vervoe to use. The system has earned $4.5 million from job platform SEEK and prominent angel investors.

The data makes the difference

"The accuracy level [of Vervoe] depends on the quality of the data," cautions Molad, "so it's still important to ask candidates the right questions. There is a risk that, if employers create their own evaluation content, the candidates might be asked to perform the wrong tasks or answer the wrong questions. This can happen if companies try to replicate their old 'interviewing style' process on Vervoe rather than putting candidates through real-world scenarios."

Eubanks agrees that having the right information is imperative.

"One of the prerequisites for having a successful A.I. implementation is data--without significant data on hand, businesses simply can't leverage what A.I. has to offer. [...] With the right approach, the insights around skills can help power smarter, A.I.-enabled practices. A few of the firms I've seen tackle this include or Catalant."

Molad says his team is working to combat this issue.

"We are currently learning which questions and tasks predict performance most accurately. We will use those learnings to feed an A.I.-powered Talent Trial creator that automatically creates evaluation content. This will give employers even more confidence in their hiring decisions because their evaluation will be based on best practice and supported by science."

Vervoe still leans on looking at every single candidate, allowing everyone to interview. This promotes a fair process. But even though tests don't take that long, if you've got tons of prospective hires, you'll need time to make the system work. If experts can find a way to create A.I. systems that can focus on quantity while successfully addressing common bias problems, Vervoe could pair with those options as a more end-stage tool to provide both efficiency and quality. If you're just getting your company off the ground or starting to scale and have fewer candidates, Vervoe might be just fine as a standalone option.

Either way, asking candidates to prove what they can do, especially in a time when people "gig it out" and can learn outside of university degrees, isn't overly intrusive. It's a legitimate way to get the ideal fit that retention, morale and profit depend on.