Juicebox App Inc. today introduced a suite of autonomous recruiting agents that help hiring teams proactively identify and engage candidates across multiple open roles simultaneously, recommending prospects, drafting outreach messages and sourcing candidates over extended periods with limited human intervention.

The new offering, called Juicebox Agents, uses artificial intelligence agents to continuously search for candidates across public online sources including GitHub, Stack Overflow, Google Scholar, Medium and community forums.

The San Francisco-based company said some early users have seen as much as a fivefold increase in recruiter efficiency and a 50% reduction in sourcing time.

The launch comes as recruiting teams struggle with a surge in job applications, partly driven by generative AI tools that make it easier for candidates to apply en masse. The company said the average candidate now submits 239% more applications than before the introduction of generative AI.

Juicebox said outbound recruiting will become increasingly important as hiring teams attempt to cut through rising application volumes and identify more specialized talent.

Map people to jobs

“With JuiceBox agents, we can actually map all people in that role,” said David Paffenholz, the company’s founder and chief executive officer. “We can look at all the machine learning-focused engineers, find the top ones and then reach out to the ones that could be a good fit for the role.”

The new product expands Juicebox’s existing AI recruiting platform, which previously required recruiters to manually search, filter and review candidate profiles. Paffenholz described the prior product as more as a “co-pilot,” whereas the new agents are designed to operate autonomously once recruiters establish hiring criteria.

The company said each agent is designed to function like an outbound recruiting partner that continuously refines searches based on recruiter feedback and evolving job requirements. Recruiters can run multiple agents simultaneously, each focused on different roles. “It’s a chat-based interface that makes suggestions for you,” Paffenholz said.

Unlike many recruiting platforms that rely primarily on scouring resumes or job boards, Juicebox aggregates information from more than 30 public data sources to build richer candidate profiles.

“Rather than purely looking at someone’s resume, we can actually look at projects that they’ve done based on their GitHub profile or essays they’ve written,” Paffenholz said. “With that additional information, we can make more targeted matches and more nuanced matches than is possible otherwise.”

Bias-free

The platform also attempts to address one of the longstanding criticisms of AI-driven recruiting systems: the potential for algorithmic bias. Paffenholz said Juicebox publishes monthly bias audits conducted by an independent third party that tests the platform against approximately 30,000 profiles across multiple bias categories.

“So far, we’ve never failed any of those audits,” he said. “We think that by being able to test the system, we can provide more consistent results and reduce bias compared to the variability of a human making those decisions.”

The agents can autonomously recommend and contact candidates, but the company said human recruiters remain involved in the process. Recruiters can review suggested outreach before messages are sent and must approve any responses drafted after a candidate engages with the company.

Paffenholz said the technology is intended to shift recruiters away from repetitive sourcing tasks and toward more relationship-oriented work. “We think that in the future, recruiters will be able to spend much more time with each candidate, build a deeper relationship and help consult a candidate on whether the role is right for them,” he said.

The product is priced by the number of active roles managed by agents. Each “agent slot” costs $200 per month and can be reassigned to different positions as hiring needs change.

Juicebox said it currently serves more than 5,000 customers and maintains a database of more than 800 million public profiles. The company has raised $116 million in funding to date from investors, including Sequoia Capital Operations LLC and DST Global L.P.

Image: Juicebox