One-Way Video Interviews
with AI Screening: How One
HR Handles
2,000 Applicants

Anyone who has ever recruited knows what that number means in practice. Every resume has to be opened, every suitable candidate invited, and every person given a response. A first pass alone — opening each resume, scanning it, making the first screening call — runs at three to five minutes per applicant, well over a hundred hours before a single real conversation happens. Add even a hundred preliminary calls, each lasting twenty to thirty minutes, and the initial funnel quietly consumes the better part of a month for one person working full-time. For a founder or a small company without a dedicated HR function, it is a particular kind of pain: hiring must happen, but there is not enough time to do it properly.

These were the pain points we kept in mind when we designed artificial intelligence features for VideoApply —  an asynchronous video interview platform. The new release takes on the most time-consuming part of hiring: helping write the vacancy with a job description, preparing interview questions, and conducting an initial assessment of candidates. For small businesses, this means one HR person or just a founder can manage the entire hiring process: no specialist department, no team expansion required. Here is what is new, and what it looks like in real HR practice.

What VideoApply Is: A One-Way Video Interview Platform

VideoApply is a one-way video interview platform — an asynchronous format in which the employer posts a vacancy and a set of questions; candidates respond by recording short videos; and the recruiter reviews them at a convenient time. No calls, no schedule coordination — the format saves substantial time.

A resume may show experience and skills on paper, but the person behind it remains invisible. An asynchronous video interview allows a first impression to form without spending a single live hour. "For me, VideoApply is a convenient way to get to know a candidate before the first call," says Nina. "Video responses let me assess experience, communication, motivation, and the ability to structure thoughts." It is especially useful for filtering out those who apply to everything almost automatically, with no real interest in the specific role. 

Simplicity has been the guiding principle — no unnecessary steps, no friction, for HR or candidate. The new AI features were built into the product with the same goal: to free people from unnecessary routine while streamlining the hiring decision-making process. (For more on getting the most out of the format, see our resources on one-way interviews.)

How To Write a Vacancy in Seconds, in Your Company’s Own Voice

Writing a vacancy used to mean starting from a blank page: spending time and effort on ideating and preparing the vacancy. Now, a short prompt and the Write with AI button are enough to have a best-in-class vacancy. A couple of words — "Senior UX/UI designer for a healthcare startup" — and the platform produces a ready-made text divided into three parts: a general description, key responsibilities, and requirements.

But the speed is not even the main advantage. The text adapts to your company's profile and tone: the same profile filled in at registration, without which a role cannot be created at all. The system keeps that profile in mind, and the listing sounds organic and natural in your company's voice. Any line can be edited manually or regenerated from scratch.

Interview Questions With a Clear Logic

The second feature handles questions, and it is the one Nina calls the most useful for her day-to-day work.

"When launching a new opening, putting together the right question set is the most demanding part for me," she says. "Now the system suggests relevant questions tied to the specific competencies the position requires." They can still be adjusted, but as a starting point they save significant preparation time. (If you prefer to build the set yourself, our guide on video interview questions to ask and avoid is a useful place to start.)

The mechanics are simple. Previously, the number of questions was fixed; now employers set between three and seven, and Generate with AI assembles them based on the job description and requirements — in a logical order, not a random one:

Opening. An icebreaker to start things off. The goal is simple: ease the tension and get the candidate talking before the harder questions begin. A cold open produces guarded answers, and the assessment ends up reflecting stress rather than skill. 

Technical. A check of professional competencies for the specific role, placed first among the substantive questions, while the candidate's attention is sharpest.

Behavioral. Soft skills and approach to work. The pairing with the technical question is deliberate: together they show both what the person can do and how they go about it.

In-depth technical. A harder test of expertise, placed after the warm-up intentionally. By this point the candidate is fully engaged and ready for it.

Motivational and cultural. How well the person fits the company. This closes the interview: after the heavier technical questions, a motivational finale lets the candidate talk about themselves rather than their credentials — and ends the conversation on a warmer note.

What matters most about this structure is not the question types, but the fact that the system keeps the interview balanced. The minimum is three questions, the maximum is seven, and the balance is enforced automatically: a second technical question will not appear until the opening and behavioral blocks are in place. The result is a well-rounded picture of the candidate, even if the HR did not design the flow themselves.

For each question, the AI also drafts the criteria against which the candidate will be assessed. If a question does not feel right, generate it, swap in your own, or cut it entirely.

One practical note: candidates get two minutes per answer. The shorter and more specific the question, the better the answers tend to be. 

AI Candidate Screening: Built-In Screening Software

The third feature is the most significant when a role attracts thousands of applications. Once a candidate finishes answering, an AI Assessment panel appears next to their video — an overall score out of ten and a summary at the top, then a question-by-question breakdown — what they said and how well it lined up with expectations. Expand the assessment criteria to see exactly what the AI was evaluating against.

Unlike resume screening tools that hunt for keywords in a document, the assessment here works from what the candidate actually said in response to role-specific questions. Back to those two thousand applications for the QA role. Reviewing each one by hand and making the initial calls, as we calculated earlier, takes the better part of a month. Pre-interview assessment changes the math entirely : candidates record on their own schedule, the AI surfaces a ranked shortlist, and the recruiter focuses on the strongest 10–15% — a few hundred people — rather than the full stack. That first pass now takes a day or two, not weeks.

One important caveat: the AI's initial assessment is a recommendation, not a verdict. It can miss things or misjudge nuance, so the final call is always yours.

The new features are in beta and free to try right now at videoapply.io.

Sorting and Responding to Candidates in One Place

This release is about AI, but it is worth mentioning the things that save time every single day. Applications are organized across three tabs: unreviewed, liked, and declined. Like and Dislike buttons under the player move candidates between them.

Responses can be sent directly from the platform to multiple candidates at once within the same tab. Messages go out by email; with names and greetings filled in automatically; templates handle the most common situations. All of this makes prompt feedback possible — without it candidates are left hanging and the company's reputation quietly takes a hit.

Transparency and Data Protection

With AI now part of the process, we paid closer attention to the legal side. Before starting the interview, candidates give double consent: to their video being recorded and stored, and to their answers being transcribed, analyzed by AI, and shared with the employer. Without both, the application cannot go through.

One point that tends to put people at ease: the AI works only with speech transcription. Video, face, and voice are not analyzed — only the content of what was said matters.

Concerns are natural. "Some people worry about privacy, some are cautious about clicking unfamiliar links," Nina notes. "These are perfectly reasonable concerns." In practice, she says, a brief explanation of why video responses are needed and who will see them is usually enough —  most candidates proceed without hesitation. Many say afterward that the format was actually more convenient than a live call: they could pick their time, record without the pressure of a real conversation, pause, and re-record.

We are upfront about what data is collected — name, email, resume, video, transcription, and AI assessment — and how long it is kept: up to 12 months. The draft application is only visible to the candidate until it is submitted, and even after submission, it can be withdrawn at any time. Everything is built to GDPR standards.

Who Will Benefit Most From the New Features

Everyone involved in hiring will feel the difference, but three groups especially so.

Founders and small teams without a dedicated HR function. When hiring falls to someone who already has a full plate, the platform handles the expert part: it suggests how to structure the role and puts together a balanced question set, even without recruitment experience. One person can run the entire selection cycle alone — from posting to decision — without adding headcount.

High-volume hiring and popular roles. Two thousand applications for one opening is not unusual: it is the norm for popular positions. Scored assessments with explanations make it easy to spot who deserves a closer look.

Teams for whom employer reputation matters. Prompt,  respectful feedback directly shapes how candidates experience — and talk about — your company.

Frequently Asked Questions

What is a one-way video interview? An asynchronous hiring format: the employer sets a fixed list of questions once, and each candidate records short video answers on their own schedule. The recruiter reviews everything later, with no calls to coordinate.

How does it differ from a live one? The candidate and recruiter are never online at the same time. Questions are set once, candidates record whenever they choose, and the recruiter reviews later. Scheduling disappears entirely, and one person can handle far more applicants than back-to-back live calls would allow.

How does AI candidate screening work? After a candidate answers every question, the platform transcribes their speech, scores each answer against role-specific criteria, and produces an overall rating and summary.  It surfaces a ranked shortlist — you make the final call.

Does the AI analyze a candidate's appearance or voice? No. The AI works only with the transcription of what was said. Appearance and tone of voice play no part — only the content of the answers matters.

Is async video interviewing GDPR-compliant? Yes. Candidates give explicit double consent before submitting — to recording and storage of their responses, and to AI transcription and analysis. Data is held for up to 12 months, and an application can be withdrawn after submission.

What Comes Next

We hold to VideoApply's core principle: simplicity without unnecessary steps, and no desire to overload either HR or candidates. Much of what is in this release came directly from client feedback — so if you feel the platform is missing something, write to us.

The advice Nina shares with colleagues also describes our philosophy: don't be afraid of automation. "The more routine that can be handed to technology, the more time is left for people." AI here does not replace the human — it gives them back the time for the work only a human can do.