Paddy Lambros spent two and a half years advising roughly 100 European startups on hiring at the London-based venture capital firm Atomico. During that time, he kept coming back to the same observation: nearly every business problem at a young company could be traced to a hiring decision. “It was either not having the right people in the role or having the wrong people,” Lambros told Fortune. A bad hire here, a stalled search there, and otherwise promising companies would grind to a halt.
That observation became the genesis of Dex, an AI-powered recruiting startup Lambros founded in early 2025. On Monday, Dex said it had raised a $5.3 million seed round led by Notion Capital, with participation from a16z Speedrun, Concept Ventures and angel investors from OpenAI and other firms.
The new funding brings the total Dex has raised to $8.4 million, including a $3.1 million pre-seed last year.
Lambros declined to comment on the startup’s valuation following the new funding.
Dex has so far focused on a single, lucrative slice of the recruitment market: AI researchers, software developers, and machine-learning and quantitative engineers. More than 15,000 engineers have signed up, and over 50 technology companies—including Lovable, ElevenLabs, Synthesia, Granola and Fyxer—are using the service.
Since beginning to charge for the product in late 2025, Dex has gone from zero to a roughly $1.8 million annualized revenue run rate, Lambros said, adding that it is “conceivable” the company could be profitable by year-end.
The startup bills itself as an “AI talent agent.” Job seekers first hold a conversation with Dex’s AI agent by either voice or text. This AI agent asks open-ended questions about a person’s experience, motivations and ambitions, then surfaces roles from a curated set of job openings, helps candidates research companies, benchmarks their compensation and prepares them for interviews. This AI talent agent is built on a mix of AI models from Google, Anthropic and OpenAI, Lambros said.
Based on the profiles these models build, Dex then uses a proprietary matching engine built from what Lambros calls “old-school machine learning” to match candidates to potential employers. If there is mutual interest from both parties, Dex passes the candidate directly to the hiring manager.
Notably, Dex does not sell its software to recruiters. “On the client side, we’re not building software for clients. We’re not building an integration for their [applicant tracking system],” Lambros said. “What they really want are great candidates.” The startup charges employers between 20% and 30% of a hired candidate’s salary, the same fee structure used by traditional executive search firms. “We only earn money if we do a good job,” Lambros said.
Kamil Mieczakowski, the partner at Notion Capital who led the round, pointed in a statement to the $856 billion global recruitment industry, where, he said, the cost to hire keeps rising even as candidates and hiring managers complain of a worsening experience. Lambros argues that AI capable of sustained back-and-forth conversation has, for the first time, made the agency function genuinely automatable—and potentially better. Unlike a human recruiter, he said, an AI agent can know hundreds of thousands of companies and speak to thousands of candidates a day.
Lambros has been in the recruiting trenches for nearly a decade. In 2016, he joined the British spatial-computing startup Improbable when it had around 50 employees, and helped grow it to 650 across the U.K., U.S. and China. He then led people and talent operations at the construction-tech startup Sensat, before joining Atomico.
Dex’s focus on AI software developers and machine learning engineers has helped the company win over a clutch of high-profile recruiting leaders. The strategy reflects what Lambros calls a hard-won lesson from earlier generations of recruiting startups such as: as platforms broaden, candidate quality tends to fall. “It’s our job to change hiring, to make it more human,” he said. “Don’t speak to 500 people. Spend really good time with these five. And then make the right choice.”
“Most recruiting tools are built by people who’ve never made a hire. Paddy has built some of the best engineering teams in the world. That deep domain experience takes years to develop, and for us it was the deciding factor,” Charlie O’Driscoll, head of technology recruitment at Monzo Bank, said in a statement.
Jonathan Durnford-Smith, head of talent at the AI video startup Synthesia, said it was “rare to find a recruiting tool built by people who have been in the very position their users are in.”
Archie Hollingsworth, co-founder of email-AI startup Fyxer, said Dex “has consistently delivered amazing candidates and made some incredibly impactful hires for us.”
Lambros is dismissive of the obvious incumbent threat—LinkedIn, which has rolled out its own AI recruiting agent. “The data set is actually quite thin,” he said of LinkedIn profiles, arguing that people are unwilling to share detailed information about their work or ambitions on a public social network.
In contrast, candidates speaking privately with the Dex agent can go into far more detail than a CV would allow. He is similarly skeptical of “arms-length” sourcing tools that scrape public data and blast out cold messages: “If the problem was cold outbound, we’d have solved this years ago,” he said.
Still, Dex is not alone in trying to build an AI platform to help tech companies recruit talented AI engineers. Rival recruitment startup Jack & Jill is also focusing on this area and raised a $20 million seed round in October. Juicebox, another competitor, raised a $30 million Series A round in September.
Dex, currently headquartered in London, plans to use its new funding to open offices in New York and San Francisco later this year.
