AI Twins Will Interview for Your Next Job: Inside TwinCafe

What Is TwinCafe?

Your next job interview might not involve you — at least not at first. On October 13, 2026, Ashutosh Garg, CEO of Eightfold AI, one of the largest AI hiring platforms in enterprise recruiting, launched TwinCafe, a side project that lets job seekers upload a résumé, LinkedIn profile, work samples and other documents to generate a shareable AI digital twin that a recruiter can interview before ever meeting the human, as Fortune first reported.

Garg envisions candidates putting a TwinCafe link at the top of their résumé or LinkedIn profile. A recruiter clicks it, asks the twin questions about the candidate's background, and gets a conversational walkthrough — before anyone books a calendar slot.

His pitch is to replace the first 20 minutes of a screening call, the résumé walk-through, with a conversation between the recruiter and the twin. "My main motivation was that a résumé is so limiting," Garg told Fortune. The twin can "fill in the blanks" on parts of a candidate's background that might otherwise raise questions — a career break, a switch between industries — so the live interview "could be a lot more focused on the human aspect versus the fact-check aspect."

He also frames the tool as a scale lever for hiring teams: "Why limit a screening to five candidates? Why not interview everyone?"

How an AI Twin Interview Actually Works

TwinCafe sits at the intersection of three AI capabilities that each matured dramatically in 2026:

• Document-grounded LLMs: The twin's knowledge comes straight from your uploaded materials — résumé, portfolio, LinkedIn — so its answers about your experience stay anchored to what you've actually done (at least in theory).
• Conversational voice: Real-time speech models make the interview feel like a conversation, not a chatbot form. This is the same wave that put voice agents in customer support this year.
• Avatar generation: Talking-head video avatars from tools like Synthesia, HeyGen and D-ID are now cheap enough and lifelike enough that a "face" for the twin is the easy part.

The candidate experience is: upload documents, review and approve what your twin knows, then share the link. The recruiter experience is: open the link, ask anything you'd ask in a screen, and skip the résumé-recitation portion of the first call entirely.

Why Now: The Avatar Stack Grew Up

TwinCafe is arriving at a moment when the underlying technology crossed a trust threshold. This week alone, HeyGen's new Voice model hit #1 on the AI voice leaderboard, virtual avatars became the fastest-growing AI video category, and always-on agents like OpenAI's Dots normalized the idea of delegating conversational work to an AI that represents you.

The economics also flipped. A 30-minute screening call costs a recruiter roughly an hour of their day once scheduling and notes are counted. A twin interview costs pennies of inference and can happen at 3 a.m. in any timezone. When Garg asks "why not interview everyone?", that's not rhetoric — it's a cost curve talking.

And the founder matters here. Eightfold AI powers talent intelligence for some of the world's largest employers. When the CEO of a major hiring platform builds a candidate-side twin tool as a side project, it's a signal about where enterprise recruiting thinks the funnel is going: asynchronous, AI-mediated, and pre-screened before any human touches it.

Recruiters Are Using AI Too — Welcome to Twin vs. Twin

Here's the part nobody in the hiring pipeline wants to say out loud: both sides are already automating. Candidates have been quietly using interview copilots that feed them answers in real time. Recruiters run AI screening agents that score recorded video interviews. TwinCafe just makes the candidate's automation explicit, visible, and — critically — consensual.

That's actually the strongest argument for the format. A covert copilot whispering answers during a live interview is deception. A labeled twin that a recruiter knowingly interviews is a new, honest contract: this AI represents the candidate, it answers from their real materials, and the human is still behind it. The endgame isn't fake candidates — it's AI agents on both sides of the table handling the fact-checking layer, reserving human time for the judgment calls.

The Honest Concerns: Trust, Bias and Verification

TwinCafe's launch claims — "the quality of the interview dramatically improves" — come from Garg himself; Fortune's piece carries no independent recruiter data yet. Beyond the marketing, real questions remain:

• Verification: What stops a twin from overstating its human's experience? A résumé can be checked; a persuasive conversational agent that "fills in the blanks" can rationalize gaps a human wouldn't survive.
• Equity: A polished twin may favor candidates comfortable with AI tooling — skewing the funnel toward the tech-savvy rather than the job-qualified.
• Disclosure norms: There's no established etiquette yet. Employers may embrace twins, ignore them, or treat a twin link as a red flag. Expect the first wave of corporate hiring policies addressing them in 2027.

Still, the direction of travel is clear. Screening is the most repetitive, lowest-judgment part of hiring — exactly the layer AI absorbs first in every industry.

The AI Interview Tools to Watch in 2026

TwinCafe is a symptom of a whole category heating up. The interview stack now spans both sides of the table:

If you're job hunting, the practical takeaway is simpler than any single tool: your professional footprint is becoming machine-readable by default. Candidates who curate what an AI can learn about them — portfolio, writing, verified work — will simply have better twins, whatever platform wins.

Frequently Asked Questions