Juicebox's AI recruiting agents go deeper than LinkedIn filters — now serving 5,000+ customers across industries

Ishan Gupta co-founded Juicebox to solve a problem that's been hiding in plain sight in recruiting: most sourcing tools pattern-match on surface signals like job title and employer, rather than asking whether a candidate can actually do the work.

Juicebox uses LLMs to go deeper. Rather than filtering for "software engineer at Google," its agents analyze GitHub activity, product work, and other publicly available signals to assess fit for a specific role. The goal is fewer outreach messages sent, to more precisely targeted candidates — which Gupta frames as a direct answer to the recruiter-spam problem plaguing senior tech executives.

“LLMs are able to truly understand what makes someone successful in a role and try to actually find people who are good at that. Instead of just pattern matching based on what company you're at or what job title you're at, we'll actually go in there, analyze your real world product, see what you're doing on GitHub, what you're doing on different platforms. We have more than 5,000 customers, and a lot of them are larger enterprises.”

Customer base and expansion

Juicebox started in technical recruiting, working with companies including Ramp and Scale. It now has more than 5,000 customers, and tech no longer represents the majority of that base. One example Gupta gives is traveling nurse placement in the Midwest, where the platform pulls from publicly available nursing registries, maps each candidate's work history across healthcare environments, and infers role fit from those signals. The same enrichment logic applies across industries.

The structural argument

Gupta's core thesis is that recruiting has always paid most for services, not software. The economic value in the industry has historically accrued to firms that can actually find and place people, not to CRM or ATS vendors. LLMs, he argues, are the first technology capable of automating the judgment that services firms sell: understanding what someone has actually built, what environments they've thrived in, and whether they'll succeed in a specific role. That's the gap Juicebox is trying to close, and Gupta believes it's the first time the software layer can capture the value that has always stayed in services.

Visibility for job seekers

For candidates trying to surface in AI-powered searches, Gupta's advice is direct: put your work product in public view. Blogs, open-source contributions, side projects. The platform enriches sparse profile data against company histories and role signals, so any publicly visible output helps. For non-technical workers without obvious artifacts, the same principle applies — the more context available across platforms, the better the inference Juicebox can draw.

Juicebox is a Thiel Fellowship company. No funding figures were disclosed in the conversation.