Today we’re excited to announce our investment in Prism, which is building an AI recruiting agent to put a dedicated talent team within reach of every founder.
As AI lowers the cost of building software, we believe more of a company’s edge will come from its team’s human judgment: what to build, which customers to serve, and how to execute. Finding the right people matters enormously, but it remains expensive and time-consuming, especially for founders without a dedicated recruiting team.
Variant invests in autonomy. Labor markets are among the world’s largest and most consequential, and access to meaningful employment is central to human agency as automation changes how we work. Prism can help more people find the right opportunities by giving founders the ability to run an effective search without building out an entire recruiting department or spending hours managing an agency, while retaining control over the key decisions that (quite literally) shape their team.
In Automate What You Can, Augment What You Can’t, we described why recruiting is a promising area for automation. Each search produces observable outcomes: whether candidates respond, progress through interviews, and ultimately get hired. Those results give a system feedback to improve future searches. As it learns which approaches lead to placements, it can apply those lessons across more searches.
But knowing whether someone was hired is different from knowing whom a company should hire. Recruiting involves subjective judgments about what a team needs, which experiences matter, and whether someone will thrive in a particular environment. Those judgments matter and are often hard for a hiring manager to articulate upfront.
Prism’s approach is to learn from that judgment before trying to automate more of the work. The company started as an AI-native recruiting agency, running searches and charging for successful placements. By doing the work itself, Prism could see which outreach generated interest, which candidates advanced, and how client feedback changed a search. Its recruiters tested approaches in live searches, giving the team a practical understanding of what software could handle and where people needed to stay involved.
That experience is informing the development of Autopilot, the recruiting agent Prism is now building for companies to configure and deploy internally. It would integrate with a company’s hiring system, surface new candidates continually, and help with outbound. Working inside the company would allow it to learn directly from the hiring team’s preferences and decisions. The aim is to reduce the coordination and review required from Prism’s own recruiters, allowing it to serve more companies without adding people to every search. We believe this is a credible path toward greater automation: start by being embedded in existing workflows, learn from the people making hiring decisions, and expand what the system can reliably do from there.
Co-founders Theo Kitsberg and Axel La Pira dropped out of Cambridge and Oxford to work on Prism, joining the YC P26 batch. They have experienced first-hand the friction of recruiting in their experience at Nightline Association (Cambridge’s mental health hotline), and at Alan, a health insurance startup. They bring the grit and raw horsepower to rethink how recruiting can be done using AI tools, and we could not be more excited to back them.