Emi Labs.

Here is a fact that sounds made up but isn't: roughly 80% of the world's workers are frontline - the people stocking shelves, driving vans, flipping burgers, staffing warehouses - and almost none of the software that decides who gets hired was built for them. Corporate recruiting spent two decades getting slick for knowledge workers. The person applying for an hourly job at a convenience store got a web form, a phone number that goes to voicemail, and silence.

Emi Labs is a bet that this asymmetry is a business, not just an injustice. The company builds AI agents that hire frontline workers, and the pitch is refreshingly literal: instead of making an applicant fight a career portal, Emi lets them apply, get screened, get scheduled, and start onboarding over the channel they already use all day, which in most of Latin America means WhatsApp. The candidate types. An AI agent answers - within seconds, at 2 a.m., in their own language, and, in principle, without ghosting them.

The origin story is unusually earnest for a piece of enterprise HR tech. Co-founders Mateo Cavasotto and Andres Arslanian met while volunteering for a microcredit NGO in Argentina, trying to work out how technology could actually move the needle for people in poverty. The answer they landed on was not a microloan app. It was hiring - the single moment where access to a job either opens or closes. They founded Emi in 2018, went through Y Combinator's Winter 2019 batch, and pointed the whole thing at the least glamorous, highest-volume corner of the labor market.

The reason this is a good business, and not just a good deed, is volume. A single retail chain in Mexico might need to hire tens of thousands of frontline workers a year, with turnover that keeps the tap permanently running. A human recruiter cannot have ten thousand genuine conversations. An AI agent can have all of them at once, and this is the pivot the entire company rests on: the thing that looks like a constraint - too many candidates, not enough recruiters - is exactly the thing software is good at. Emi says it has now run more than 12 million candidate conversations. That is not a marketing number so much as a description of what the product is.

What Emi sells to the enterprise is boringly practical. The AI agents attract candidates from job boards, social media, QR codes on a store window, and referral links; screen and rank them in real time against the client's criteria; schedule interviews automatically; and then generate the employee record in the client's HRIS - the system of record - in minutes rather than days. The company claims the automation cuts onboarding errors by more than 80%, which matters because a mis-keyed document on day one is how an hourly worker ends up unpaid on payday. Emi also reports a roughly 50% reduction in time-to-hire and a 90% vacancy fill rate for its customers. Treat vendor metrics with the usual salt, but the direction is the point: speed is a feature a candidate can physically feel, and in frontline work, the fastest offer often wins the worker.

The customer list is where the story stops sounding like a seed-stage dream and starts sounding like infrastructure. Emi's agents screen candidates for Walmart, OXXO, Coca-Cola FEMSA, Mercado Libre, Home Depot, Heineken Mexico, Danone, Liverpool, Alsea, Burger King, and Domino's Pizza, among 50-plus enterprises. The company's most striking claim is a market-share one: that roughly one in ten frontline hires in Mexico now passes through its platform. If that holds, Emi has quietly become plumbing - and plumbing is a much better business to be in than features, because nobody rips out the pipes.

The money followed the logic. Emi raised an $11 million Series A in February 2022, led by Merus Capital and Khosla Ventures, with Flexport Ventures, FiDi Ventures, Lorimer Ventures, and PrimeSet also in. Add the earlier seed capital that came after Y Combinator and the total sits around $13 million - modest by the standards of AI hype, which is itself telling. This is a company that reported roughly $10.5 million in revenue in 2023 off a team that had grown past 100 people. It is trying to be a real business, not a valuation.

What makes Emi genuinely interesting - the part worth watching - is that it built for the overlooked user and let the enterprise budget follow. The frontline worker is the customer nobody optimized for, and Emi made them the entire point of the product. The bet is that if you design the candidate experience to be fast, human, and hard to ghost, the employers who need those candidates will pay to be on the other end of the conversation. So far, the employers are paying.

None of this is finished. Conversational AI at scale is a moving target, competitors like Paradox, Sense, and Fountain are chasing the same high-volume prize, and "an AI agent screened you" is a sentence that will earn scrutiny as fast as it earns hires. But Emi has a defensible answer to the oldest problem in hiring - too many people, too little time - and it aimed that answer at the workers most software forgot. That is a narrow door into an enormous room.