Skills-first hiring in the era of AI

Shay David: In the early days of the company, we would talk with people about the power of AI, and sometimes they would roll their eyes and say, “Yeah, but that’s years away.” ChatGPT was the marketing campaign we could never afford.

Shana Ting Lipton: That’s Shay David. And today, we hear the story of how his talent management startup, Retrain.ai, is using artificial intelligence to propel skills-first hiring into the future. I’m your host, Shana Ting Lipton, with PwC, and this is Voices in Tech, from our management publication, strategy+business, bringing you stories from the intersection of technological acceleration and human innovation.

Over the past few years, a series of developments have contributed to a radical evolution in talent management. First, it was a hiring boom in the late 2010s, followed by covid-related worker shortages and new ways of working. There’s also been a surge in popularity of online learning and microcredentials, and, of course, the present and looming impact of AI as both a challenger to human-held jobs and a facilitator of human employment through sophisticated skills-matching tools.

Launched in 2020, Israeli talent management platform Retrain.ai analyzes real-time market data alongside enterprise data to skills-match talent and jobs, and predict imminent skills gaps and jobs of the future. CEO Shay David.

David: So, in my previous role, I was the cofounder and for many years the chief revenue officer of a company called Kaltura. Kaltura was an enterprise-video platform selling video-platform solutions to many of the world’s largest organizations, and one of the killer use cases for Kaltura was and remains enterprise training and learning. Organizations are faced with an ever-changing market with ever-changing demands, and they need to develop skills. And video was just a very effective way of developing those skills. But the problem of the skills gap is much larger than video. So, over many years and many conversations with customers, one thing became pretty apparent, which is organizations actually do not necessarily understand the skills that they need and the capacity or the skills that employees have.

Lipton: So David joined forces with a tech investor, Isabelle Bichler[-Eliasaf], and a chief data officer, Avi Simon.

David: Isabelle and I have known each other from the New York tech ecosystem. Isabelle was an investor and I was working with startups, and she had an interesting portfolio. But we always wanted to look for the big idea that we can work on together. So, when Retrain came along, it was obvious for us, we wanted to do this together. And Isabelle and I knew that we needed a deep technical bench and we met Avi. So it became apparent through the statistics and through many conversations that dual tragedy—on the one hand, businesses cannot find employees; on the other hand, many people cannot find work. The reason for that is the skills gap problem. We realized that the number one limiting factor for growth was [lack of] access to skilled talent.

Lipton: A challenge that was highlighted in PwC’s last CEO survey [published in 2023], which found that 52% of CEOs believe that labor and skills shortages will impact their industries’ profitability in the next ten years.

David: So, we realized that what the market really needs is a global skills framework. It’s a dynamic role catalog that allows organizations to understand skills on a global scale—allows them to understand which jobs exist, what are the tasks and skills required for the jobs, and then be able to place people in training pathways on the same skills framework, so that in a lingua franca, or in a common tongue, we can understand people, jobs, and training pathways. Imagine a bank that is changing its business, closing down branches on Main Street, moving to online banking. Imagine in a situation like that, what does it mean to be a project manager at a bank like that? In the physical world, project management may be thinking about physical customer experiences or building projects or something like that. In a digital environment, we’re talking about digital-only experiences. So the title project manager remains the same, but the capacity and the skills and the tasks completely change.

Lipton: Just as David, Bichler, and Simon were poised to launch the company in 2020, the unexpected happened.

David: So, in early 2020, covid hit and dramatically accelerated the timeline for the future of work. Many organizations had to change the way that they work. All of a sudden, their employees worked at home. They needed to do things differently, and many organizations found themselves in a bind. If before, they had a problem with skilled talent, now that problem only got bigger. So, the hearts and minds were really ready for a system that’s going to allow them to understand the gaps and to be able to help organizations take decisions around core HR processes, using the language of skills. We decided that covid was not going to stop us, and we were able to raise seed funding, hire our first employees, and basically get off the ground without ever meeting in person.

Lipton: Something else was going on in 2020: a global startup boom, which saw Israel, sometimes called “Silicon Wadi” (Hebrew for Silicon Valley), emerge on top. In fact, the Israeli government’s innovation authority had begun to lay the groundwork for the startup ecosystem in the ’90s through early promotion of tech incubators and venture capital.

David: So, we decided to build core engineering in Tel Aviv. Tel Aviv has a very striving tech ecosystem and lots of engineering talent. We also tapped into some talent in Palestine. We have an Israeli–Palestinian collaboration with a small team of developers in Nablus. And later, as we expanded, we’ve actually expanded to a small Polish team in Warsaw. While the engineering team is all here on this side of the ocean, our go-to-market team and sales and marketing teams are all in New York City. One of our first customers was the Israeli Department of Labor. The main challenge they were facing is getting a real-time understanding of the labor market. In many countries, Israel included, there’s a Bureau of Labor Statistics or Central Statistics Agency; and they use typical statistical tools, employer surveys, and many kinds of after-the-fact tools to be able to understand the labor market. But if what you really want to do, particularly in the context of post-covid recovery, is direct millions of people into vocational training or help employees find better jobs or help employers find talent, then you really need the data to be fresh. And one of the things that we were able to do is, using the power of AI, to scan dramatically larger data sources so that we can give the government an analyzed version of a real-time map of the labor market as it was changing from month to month and day to day.

Lipton: The platform has plugins and operates in tandem with existing systems of record, like Workday, to add a layer of skills intelligence. So how does it actually work?

David: It’s able to scan many different information sources, such as job boards or LinkedIn. And then, it’s using an AI technology called NLP (natural-language processing) to be able to understand text—plain English text or plain Hebrew text or any other language—and to be able to capture the skills information from that. So, in essence, we developed two main technologies. One is a big knowledge graph that understands the relationship between roles and tasks and skills. And the second is a language engine that is able to take normal text and be able to map that into that graph. Once you have both of those components, then you can ask a lot of questions about the data and understand it in a dramatically more granular and actionable way. You can also place that data on a map in a geographic information system. So now you can gain geographic insights as well.

Lipton: In 2021, as the startup was expanding its total funding to US$20 million, record numbers of people around the globe were leaving the workforce. While the “great resignation” may largely be over, PwC’s 2023 Global Workforce Hopes and Fears Survey tells a story of continued employee restlessness, along with skills inequities. Among employees who took on second jobs, 36% did so because they wanted to learn new skills. Bhushan Sethi, Strategy&, principal with PwC US.

Bhushan Sethi: Yeah, it’s confusing businesspeople. It’s confusing economists. One in four people in our survey are saying that they’re going to be looking for other jobs. They’re also saying that they’re going to be demanding pay rises and promotion at a greater level than last year’s survey. It just highlights the real tale of two workers. We see specialist workers, those that have specific qualifications and experience, having a much more fulfilling experience, being much more open to thinking about new skills, and having greater job satisfaction and opportunity. And we see people that have less specialist skills kind of struggling—struggling kind of with well-being, struggling with living paycheck to paycheck. That’s a big challenge for organizations and society.

David: So, right after covid, one of the most interesting phenomena we observed was that while employers were having a hard time finding talent, many people, millions of people, were having a hard time finding jobs. People that could not develop 21st-century skills were becoming either unemployed or the risk was that they would become completely unemployable, or underemployed. These are people that don’t have skills that the market can value. The solution in our minds is to help them tap into the same skills platforms so that they can understand what are the skills that are 21st-century skills, and how could they find training pathways that are going to give them the path back to full employability?

Lipton: 35% of employees PwC surveyed also said they have skills that aren’t clear from their qualifications, job history, and job titles, with 27% saying employers focus too much on job histories and not enough on skills. This is a clarion call for skills-first hiring, an increasingly popular approach to talent management that puts more emphasis on skills than on degrees, job histories, or job titles.

David: In the past, for many jobs you needed to have academic credentials like a BA or an MA, or maybe even a PhD. Today, employers understand that when you’re getting one of those academic degrees, essentially you’re getting a basket of skills. But in a rapidly changing market, what employers want are specific skills. Let’s take an example from programming. Maybe you have a degree in computer programming, but if your employer wants you to program in a specific language which you don’t know, then you might be even at a disadvantage versus somebody that learns programming in a specific language and two courses on Udemy. But now they have the skills to get a specific job done. So, microcredentialing and the capacity to do specific types of work and specific tasks is critical.

Lipton: Another generational shift shook things up in late 2022. Also a game changer for skills and the workforce.

Woman’s voice: ChatGPT.

Second woman’s voice: ChatGPT.

Man’s voice: ChatGPT.

David: ChatGPT was the marketing campaign we could never afford, because in the early days of the company, we would talk with people about the power of AI, and sometimes they would roll their eyes and say, “Yeah, but that’s years away.” Nobody that ever chatted with ChatGPT is going to say that this is years away, because they understood the power of very large language models. And that really helped us bring to the fore this idea that an AI agent, an HR copilot if you will, is literally around the corner, and in fact there’s some sort of race of building that HR copilot, which is exactly what we’re building.

Lipton: So, what are the core skills required for what the World Economic Forum calls the jobs of tomorrow?

David: Soft skills are mostly focused around capacities for collaboration and problem solving. Whether that collaboration is between humans or whether that collaboration is human and machine on one hand. And hard skills is [sic] very much focused on the application of technology into the workplace. We see that technology in that sense is pervading both soft skills and hard skills. Pretty much every profession in the world—whether that’s warehouse employers or whether that’s programming—requires the use of tools. And we can capture that idea with this notion of digital literacy, the capacity to use digital tools for whatever job is at hand. This is a really strong driver in the changing labor market.

Lipton: Bhushan Sethi.

Sethi: There’s a thirst for learning around AI, and there isn’t the fear that we see in terms of headlines around “artificial intelligence is coming for our jobs.” So that’s a good thing for business. Again, we need to kind of make sure that we are enabling and equipping people with the right skills and, obviously, the right governance and controls around AI and datasets in business.

David: When you think about the power of AI, you have to make sure that you’re developing AI in a responsible way that is going to help humans become better versions of themselves rather than amplify their biased tendencies. So, the idea of responsible AI is to put forward a set of design principles that are going to make the AI system controlled, that are going to make it transparent, that are going to make it auditable, in a way that you can use AI in order to help an organization take better decisions without amplifying bias. So, we believe that AI can really be a force for good. AI is not only a technology that’s going to eliminate or change jobs, but AI could really be the technology that helps humans maximize their potential.

Lipton: Thanks for listening and stay tuned for more episodes of Voices in Tech, brought to you by PwC’s strategy+business.