Moonhub’s AI chatbot, trained on 1 billion candidate profiles, helps hiring managers at AI unicorns like Anthropic and Inflection find and recruit employees.
In August, Tony Zavala was hired as an AI language specialist to improve the performance of Pi, a conversational chatbot developed by unicorn startup, Inflection. But Zavala, who previously worked as a computational linguist at LinkedIn, wasn’t looking for a job. He was discovered and recruited thanks to an AI agent built by San Francisco-based startup Moonhub.
A human recruiter had provided the AI with specifications like “experience working with ChatGPT-like products,” and “technical experience at a top tech company” to find Zavala’s profile. Before Moonhub’s human recruiter reached out, Zavala said he wasn’t “specifically familiar” with Inflection or the job opening and had only heard the company’s name in passing.
“We help companies find those we call ‘the hidden talent.’ Oftentimes [candidates] are not actively interviewing, but they may be passively interested in a conversation,” said Moonhub founder and CEO Nancy Xu.
Since launching in January, Moonhub has onboarded more than 100 customers including billion-dollar startups like Inflection and Anthropic as well as other startups like Atomic AI, Hippocratic AI and You.com Xu told Forbes. The company announced Wednesday that it raised $10 million in seed funding from Khosla Ventures and GV with participation of Time Ventures, Day One Ventures and angel investors like Susan Wojcicki (former YouTube CEO), Mike Volpi (Index Ventures general partner) and others. Within a year since launch, the startup has booked more than $1 million in revenue, Xu said, but the purpose of the capital infusion is to improve the AI product and to expand the team.
“At Anthropic, as well as at Inflection, we were one of the earliest talent partners that they worked with,” Xu told Forbes. “Actually, in the case of Anthropic, we worked with them before they even had their own in-house recruiter.”
Recruiters can query Moonhub’s conversational chatbot to find hundreds of relevant candidates by feeding it contextual information like the job description and company information. They can also inject prompts like “show me people in the Bay Area from top Ivy League Schools” and find people working at similar companies or companies that use similar technologies. Companies use Moonhub’s recruiting services to hire for white-collar jobs like data scientists, software engineers and marketing managers but Moonhub’s clients aren’t only limited to tech companies, Xu said–they also include elder care company LifeWorx, electric boat manufacturer Navier and parking space provider Diamond Parking.
Moonhub’s AI recruiter is trained on a database of more than 1 billion public profiles sourced from a range of sites including LinkedIn, Upwork, GitHub, Google Scholar, Overflow, StackOverflow and Twitter. The company used its in-house large language models as well as those developed by OpenAI, Cohere and Anthropic to create its own conversational recruiting agent. Of note, Xu said, is its patent-pending “custom retrieval augmented generation framework,” a technique that can improve the quality of the chatbot’s responses than a regular LLM by quickly processing additional information like recruiting knowledge from experts in the field.
As of now, Moonhub’s human recruiters reach out to potential candidates, schedule interviews and manage the recruitment process, but soon AI will handle everything apart from interviews, she said. But using AI for finding and screening candidates can introduce several biases in the hiring process and can potentially exclude and discriminate against candidates from certain backgrounds. Xu said her startup is trying to make the AI recruiter more fair by allowing employers to boost profiles of diverse candidates and diversifying the data its models are trained on. She said Moonhub is a modern AI alternative to startups who outsource recruiting duties to external firms – a costly and time-intensive experience. “Many companies that are not AI native companies themselves are coming to Moonhub as a way to become an AI native company.” said Xu, who also runs her own VC firm, Xu Ventures.
Before starting Moonhub, Xu left a PhD in computer science at Stanford, where she worked on foundational models to build AI agents for various applications including customer support at the Stanford AI Lab. She also ran a recruiting firm where she learned about the difficulties of finding quality candidates in a short span of time, which eventually led her to create the startup.
“For many people the magic moment when they use Moonhub is when they realize that they can sit back on a Friday night, watch Netflix, chat with an AI, and in five minutes, they discover 50 candidates that they really like,” Xu said.
