Our Vision

The universal first-round interviewer

Companies receive billions of applications a year and speak to almost none of them. Qualified people are filtered out by a resume scan, recruiters drown in volume, and candidates are ghosted by both. We think every applicant deserves a real conversation — and every company deserves to find the best person rather than the best resume.

The first round is the broken part

Almost everything wrong with hiring happens in the gap between applying and speaking to a human. It is where good candidates are discarded for keyword reasons, where the process stalls for weeks, and where an employer's brand quietly takes its damage. It is also the only stage where the work is genuinely repetitive: the same questions, asked of everyone, scored the same way.

That combination — high volume, high stakes, low variety — is exactly what should be automated, and exactly what has been automated worst. The first wave of automation removed the conversation entirely and left a keyword filter. The second replaced it with a voice on a phone line or a chat box, which is why candidates started walking out.

We think the answer is not less automation. It is an automated first round that is worth having: a real conversation, on the candidate's schedule, that adapts to their answers, treats every applicant to the same standard, and hands the hiring team something far richer than a resume.

Why now

Technology

Realism became real-time

Photorealistic faces and conversational language models both crossed their thresholds, and then the two became cheap enough to run together inside the half-second a conversation allows. Interviewing was waiting on that intersection.

Market

Volume broke the old model

Applications per opening keep climbing while recruiting teams shrink. The response has been keyword filters that reject people no one ever spoke to — a screening layer nobody trusts, on either side of the table.

Sentiment

The backlash already happened

Candidates are not refusing AI interviews; they are refusing bad ones. Nearly four in ten have walked away from a hiring process because of one, and most were never told an AI was evaluating them. Meanwhile, given the choice between an AI interview now and a human interview later, most take the interview.

Sentiment figures: Greenhouse Candidate AI Report, 2026 and a University of Chicago field study, 2025.

What we've built

Presence is the whole bet. A voice on a phone line is a feature that any hiring product can add in a quarter; a face that holds together in real time, through every pause and interruption of a live conversation, is an infrastructure problem that takes far longer than that.

A face rendered live

Ashley is generated frame by frame while she speaks, in step with words her language model produced a moment earlier. She is not a video avatar reading a script, and she is not assembled from pre-recorded clips — which is what makes her able to be interrupted, to pause, and to react.

An inference stack we own

The rendering pipeline is built and operated in house rather than rented from a video-avatar API. It is the slowest thing we built and the hardest to copy, and it is why the rest of the category ships a phone call or a chat window.

Trust designed in, not retrofitted

Ashley discloses that she is an AI at the start of every conversation. Scoring reads the transcript — the words — against the employer's rubric; appearance and accent are never evaluated. A person on the hiring team makes every decision. In a category defined by candidate backlash, that posture is a product feature rather than a legal afterthought.

Where this goes

The near term is the first round for roles where presence matters and volume hurts — the screens a team would run by hand if it had infinite hours. Ashley is used by teams at the U.S. Air Force Academy and the University of Florida's Warrington College of Business.

Further out, an interviewer that every applicant actually meets produces something no resume database has: a structured, consistent record of how people answer questions about the work, at a scale human interviewing has never reached. That is the foundation for understanding what a strong candidate for a given role actually sounds like — and for a hiring process that finally judges the conversation rather than the paperwork.

Investors & partners

If you are building in hiring, evaluating the category, or want to see the live interviewer for yourself, we would like to hear from you.

anthony@tryashley.ai

← Back to Ashley.AI