Wanted Lab looks beyond hiring with AI transformation push — 'We solved our own problems first'

In-house automation experience commercialized as Ennoia

Frontline staff build and deploy their own agents

AX sales hit 10.3 billion won; company targets 50-50 split with recruitment this year

Lee Bok-gi, CEO of Wanted Lab, at the company's headquarters in Songpa-gu, Seoul. [Provided by Wanted Lab]

"A company that solves its own problems is a true AI transformation company. We need to create an environment where everyone can become a developer and everyone can become a builder."

That was how Lee Bok-gi, CEO of Wanted Lab, explained why a recruitment platform company is expanding into AI transformation, or AX, services. On the surface, hiring and enterprise AI solutions may seem unrelated. But Wanted Lab has spent more than a decade training data to solve the problem of matching people with companies. With the rise of generative AI, the company says it extended that accumulated data expertise and problem-solving approach beyond recruitment to enterprise operations more broadly.

Lee spoke with reporters Thursday at Wanted Lab's headquarters in Songpa-gu, Seoul. "Wanted Lab is fundamentally a company that handles data, trains it, and uses it to solve problems," he said. "The direction of 'AX beyond HR' came from asking whether we could apply the experience of solving matching problems with data in HR to a wider range of customer challenges."

Wanted Lab's AX business was not built overnight to chase a new market. It grew organically from late 2022, as generative AI spread and employees began automating repetitive tasks and building their own agents and services. Demand from people wanting to learn how to solve problems with AI gave rise to "AX Champion," a hands-on training program. The need for a safe environment to build and deploy services became Ennoia, an enterprise AI agent platform. Requests to have solutions built on their behalf expanded into the company's Gigs business.

Ennoia gives frontline employees the tools to build and deploy AI agents tailored to their own work, along with security and access-management controls. Users describe what they need in plain language, and the platform automatically assembles verified functional building blocks. The system is deployed on-premises, keeping sensitive data from leaving the company's internal servers. Data access can also be configured by department, seniority level or individual user.

"We are not handing people a finished tool — we are laying the groundwork so that frontline staff can build things themselves," Lee said. "In the past, you had to submit a request and wait for a change. Now, what needs to change today can be changed today." He added that systems built to today's technology standards could already be outdated by the time they launch a year later. "We need to transplant startup speed so that frontline staff can build things themselves," he said.

Wanted Lab validated this approach internally before taking it to market. More than 150 AI agents built by employees now run across every division — human resources, general affairs, sales, marketing and finance. Examples include a general-affairs agent that automates visitor verification and door access, a system that processes book loans by scanning a barcode, a service that lets staff query internal data in plain language, and an agent that handles routine HR inquiries.

Through the AX Champion program, frontline employees bring their most frustrating work tasks and spend two to four weeks building their own agents. "When you look at the people who build the most in-house, there are actually more non-developers than developers," Lee said. Products built by employees that show potential for external use are released through "Wanted Lab," an internal incubator. Services that have emerged from it include one that maps a user's career profile based on their resume and another that surfaces nearby job listings on a map.

Development speed has also improved sharply. Projects that once required eight people working four to eight weeks are now being completed by three people in one to two weeks. As service development accelerated, governance features — controlling which agents are deployed and who can access internal data — were built directly into Ennoia.

Ennoia, proven internally, is now being rolled out to large corporations and public institutions. About 200 agents are currently running for roughly 16,000 users, and at one major automotive parts manufacturer, more than 12,000 employees are using about 30 agents.

That company has turned scattered design and technical knowledge into searchable assets for use in product development. It has also connected parts data with overseas quote-request data to identify business opportunities, and automated repetitive tasks in general affairs and compliance.

The AX Champion program has logged about 3,000 cumulative participants. Wanted Lab runs the program for corporations, public institutions and universities, using it to identify operational pain points and then channel participants toward adopting Ennoia or engaging the Gigs business.

Wanted Lab's AX revenue grew from 2.3 billion won ($1.48 million) in 2021 to 10.3 billion won in 2025. Its share of total sales rose from 7 percent to 27 percent over the same period. The company aims to bring the AX business to a 50-50 revenue split with its recruitment business by the fourth quarter of this year.

"I don't want to stop at simply putting the word 'AI' out front," Lee said. "What matters is generating real results through actual business and creating a virtuous cycle that feeds back into investment. Ultimately, the companies that will stand out are those that have used AI to solve their own organization's problems themselves."

rim@heraldcorp.com