The elevator pitch sounds almost quaint in an era of chatbots and chat windows: What if candidates actually spoke to an AI during their first screening, rather than filling out forms or messaging back and forth?
That's the wager Ezra AI Labs is making. Fresh from Andreessen Horowitz's Speedrun accelerator program, the Lafayette, California-based startup has built a voice-first interviewing platform already integrated with Greenhouse and Ashby—two heavyweights in applicant tracking software. Rather than text-based automation, Ezra conducts phone interviews, evaluates responses, and delivers rankings and transcripts back to recruiters. It's structured screening, just conducted by voice instead of the usual digital questionnaire.
Ophir Samson, the founder, isn't new to voice AI. This is his second venture in the space, following stints at GM and Aurora, plus earlier research in mathematics at MIT. He holds a PhD from Imperial College London and an MBA from Stanford—a background that blends technical depth with business school polish. In promotional materials, Ezra positions itself as a "fairness-first" solution, a phrase that carries particular weight given ongoing debates about algorithmic bias in hiring.
How It Works (and What It Costs)
Ezra's platform runs phone-based screening interviews customized for specific roles. Once a candidate completes the call, recruiters receive transcripts, summaries, ranked evaluations, and what the company describes as "cheat detection"—flags for AI-generated script reading or other irregularities. The results flow directly into a company's existing ATS. Greenhouse has published integration documentation; Samson announced the Ashby partnership on LinkedIn.
Pricing follows a freemium structure. Three interviews per month come free. A $200-per-month Starter tier covers fifty interviews, according to the company's pricing page. Whether enterprises are paying substantially more for higher volumes remains unclear—Ezra hasn't disclosed customer contract details.
The Team Behind the Platform
Samson assembled a team with startup and brand-name tech credentials. Melissa McNutt, the chief marketing officer, comes from Samsung and Magic Leap. Marc Yasuda, leading commercial efforts, previously worked at LinkedIn and Headspace. Anna Matalon, heading product, is a Y Combinator alum who founded a fintech startup and later worked at Scale AI.
According to the Andreessen Horowitz Speedrun Cohort 005 directory, the company had four employees during the accelerator program. A recent job posting claims the company has completed "two rounds of funding" and names an eclectic mix of backers: Andreessen Horowitz, Penny Jar Capital (Steph Curry's investment fund), LMNT Ventures, and angels from NEA and Sequoia. No funding amounts have been disclosed publicly, and a search of SEC filings turned up nothing. For a company actively hiring founding-level engineers and sales staff—listings appeared on Wellfound and similar boards—keeping the exact figures quiet isn't unusual at this stage.
A Crowded, Consolidating Market
Ezra enters a recruiting automation landscape already thick with competitors, and one that's consolidating fast. Workday announced plans to acquire Paradox—the company behind Olivia, a widely deployed recruiting chatbot—last August, with the deal expected to close sometime in the current fiscal year. HireVue absorbed Modern Hire back in 2023 and has been expanding its AI interview offerings since. Then there are the incumbents with long-standing interview tools, video assessments, and skills testing platforms.
Still, adoption is clearly accelerating. Employ Inc.'s most recent Recruiter Nation Report found that 65% of recruiters now use AI to enhance their recruiting technology. The conversational AI market overall—spanning customer service, sales automation, and recruiting—hit $12.24 billion in 2024, according to Fortune Business Insights, with projections reaching $61.69 billion by 2032. Recruiting is only a slice of that broader pie, but it's a slice investors seem eager to claim.
Ezra's website displays logos from Zapier, Ello, Dandy, Zendrive, and Clover Health, alongside testimonials. Whether those represent paid pilots, full deployments, or marketing partnerships wasn't independently verified.
The Fairness Question
Voice-based hiring tools face scrutiny that text-based ones sometimes sidestep. Accent recognition, dialect variation, speech impediments—these introduce variables that can advantage native speakers or penalize candidates from specific backgrounds. New York City's Local Law 144, in effect since mid-2023, requires annual bias audits for automated employment decision tools used within the city. Research continues to surface concerns about equity in voice AI systems, from ASR (automatic speech recognition) accuracy gaps to the potential for proxy discrimination.
Ezra emphasizes "fairness" and structured evaluation as differentiators. The website talks about reducing bias through consistent questioning, but the company hasn't published independent audit results or third-party fairness evaluations. That may come—regulations in New York and potentially other jurisdictions will likely force the issue for any recruiting AI that gains traction. For now, it's a pitch more than a proven claim.
What Comes Next
The company appears to be transitioning from accelerator momentum into actual go-to-market execution. Job postings reference enterprise pilots underway, investor backing in hand, and founding-team roles still open—typical signals of a startup moving from build mode into sell mode.
Whether voice-first screening resonates with enterprise buyers remains to be seen. Recruiters already juggle video interviews, skills assessments, background checks, and reference tools. Adding another layer—even one that promises efficiency—means convincing talent teams that phone-based AI interviews deliver something their current stack doesn't. And that they can defend the approach if a candidate, or a regulator, starts asking hard questions about how the algorithm makes its calls.
For Samson and his backers, the bet is that voice unlocks something text can't: a more natural, accessible, perhaps even more equitable first conversation. The market will decide if they're right.
