The pitch sounds deceptively simple: What if job-hunting engineers could just talk to an AI agent about what they actually want, instead of polishing a resume and waiting for their LinkedIn inbox to fill with templated messages from recruiters who clearly haven't read their profile?
That's the premise behind Dex, a London-based startup that's raised $5.3 million to build what it calls a "talent agent" for technical candidates. The seed round, led by Notion Capital with backing from a16z Speedrun and Concept Ventures, brings the company's total funding to around $8.4 million as of late April 2026. It's a bet that engineers are tired of the spray-and-pray tactics that dominate hiring—and that they'll trust a conversational AI more than they trust the standard recruiting machinery.
The model is unusual. Candidates don't fill out forms or upload a CV. Instead, they have a conversation with Dex's AI—either speaking aloud or typing, depending on their preference—about their experience, what motivates them, and what kind of role they're looking for. The agent asks follow-up questions, builds a profile that goes beyond bullet points, and offers compensation benchmarks and company research along the way. When there's a match with a hiring company, Dex makes a direct introduction to the hiring manager. No applicant tracking system. No software subscription. Just a traditional success fee—20 to 30 percent of first-year salary—if the candidate gets hired.
It's a model borrowed from executive search, applied to a generation of engineers who expect their tools to be conversational and their interactions to feel less transactional. Whether it works at scale remains an open question.
The Agent as Middleman
Paddy Lambros, Dex's co-founder and CEO, comes from the talent side of venture capital. He previously advised portfolio companies at Atomico and led people teams at Sensat and Improbable, two UK tech firms that navigated rapid hiring cycles. His co-founder, Harry Uglow, worked in engineering at Atomico. The two launched Dex in February 2025, and by April 2026 the platform had attracted more than 15,000 engineers and over 50 companies as clients, according to Fortune. Early customers lean heavily toward AI-native startups—ElevenLabs, Synthesia, Lovable—the kinds of companies comfortable with agent-based tools and chronically desperate for engineering talent.
The technical architecture relies on foundation models from Google, Anthropic, and OpenAI to power the conversational layer, Lambros told Fortune. The matching engine, he said, uses "old-school machine learning" trained on the rich profiles those conversations generate. It's a hybrid approach: cutting-edge language models on the front end, more traditional algorithms doing the actual pairing work in the background.
The multimodal design—candidates choose voice or text—marks a subtle evolution. When Dex first surfaced in coverage by TechCrunch in 2025, it was framed as an "AI voice talent agent." The expansion to text appears aimed at candidates who prefer writing or who simply can't talk aloud in an open-plan office or during a commute. It's a practical acknowledgment that even in an age of voice interfaces, not everyone wants to speak to their phone.
Revenue, Not SaaS
Dex doesn't sell software to HR departments. There's no enterprise contract, no seats to license, no integration with Workday or Greenhouse. The business model is straightforward: if a hire happens, Dex gets paid. If not, it doesn't. The company began charging fees in late 2025 and reported approximately $1.8 million in annual recurring revenue as of April 2026—a figure that, if accurate, suggests steady but not explosive traction. Lambros hinted that profitability might be within reach by the end of the year, though that timeline will depend on how aggressively the company invests in expansion.
The approach sidesteps the crowded market for recruiting software, where LinkedIn, Eightfold AI, and a constellation of HR tech vendors compete for enterprise budgets. Instead, Dex competes with headhunters and agency recruiters—a fragmented, high-margin world where success often depends on relationships, networks, and the ability to surface candidates that others miss. The promise is that an AI agent can do that work at scale, surfacing talent that traditional search overlooks while avoiding the inbox spam that alienates both candidates and hiring managers.
Whether companies will trust an algorithm to make those introductions—and whether engineers will engage honestly with a bot—is the real test.
A Wave of Agent-Based Recruiting
Dex is hardly alone in trying to automate parts of the hiring process with AI agents. The first half of 2026 has seen a flurry of product launches. Eightfold AI, a well-funded enterprise player, expanded its "Talent Agents" across the interview lifecycle in April. Joveo launched an AI interviewer with voice avatars in March. uRecruits debuted autonomous pre-screening and scheduling agents—complete with voice, video, and text options—at the end of April. And LinkedIn's Hiring Assistant, an agentic tool for recruiters, became globally available the previous September.
The timing isn't coincidental. Foundation models have gotten good enough to handle nuanced, multi-turn conversations, and the engineering talent shortage hasn't eased. Companies are willing to experiment with tools that promise efficiency gains, and candidates—particularly younger engineers—seem less put off by the idea of talking to an AI than previous generations might have been.
Dex's focus on technical roles and AI-forward companies may offer some insulation from the noise. The early customer base skews toward startups that are both comfortable with agent-based tools and operating in fiercely competitive hiring markets. But that niche won't stay protected forever.
Going West
Dex plans to open offices in New York and San Francisco later this year, according to Fortune. The move is almost obligatory for a recruiting platform with ambitions beyond Europe—Silicon Valley and New York remain the deepest talent pools for the kinds of engineers Dex is targeting. The London headquarters gives the company a foothold in a market with its own density of AI startups and scaleups, but the real growth opportunity lies across the Atlantic.
Notion Capital, in its investment thesis, cited the global recruitment industry's $856 billion valuation—a figure large enough to accommodate multiple business models and plenty of competitive jostling. Whether Dex can carve out a defensible position will depend less on the sophistication of its AI and more on execution: how well it delivers on the promise that matters most in recruiting, which is making the right introductions at the right time.
For now, the startup is making a wager that engineers want something different from the hiring process—something that feels less like being processed and more like being understood. Whether an AI agent can deliver on that feeling, at scale and at a profit, is the question the next eighteen months will answer.
