Automated AI interviews: real questions, scores backed by evidence

The most people-intensive step in recruitment isn't posting the job, it's the first-round interview. Your leader has to meet dozens of candidates, ask the same few questions over and over, and then most get rejected after five minutes. automated AI interviews was built to carry exactly this round: the machine talks with candidates through real work scenarios and scores objectively with evidence, so your people only meet those truly worth meeting. In this article we explain how it works and why MONA believes in this approach.

What an automated AI interview actually does

Many people picture a machine reading out fixed questions and making candidates answer multiple choice. That's not it. A properly built AI interview system plays the conversation partner: it sets up a scenario, listens to the answer, then asks a follow-up based on what it just heard.

For a customer care role, for example, the machine plays a customer annoyed about a late delivery. Depending on how the candidate handles it, the machine pushes a harder scenario, just like a real working session. This dynamic questioning gives a much closer picture than reading a polished CV. It is the natural next step after you have already AI CV screening at the application stage.

A big difference: the machine asks everyone the same way, with no bias for looks or accent, and it doesn't get tired after the twentieth person. Every candidate is judged against the same yardstick.

Why scenario-based interviews are more trustworthy than theory questions

A question like “what do you think your strengths are” is nearly useless, because everyone memorizes a nice answer. Smooth talkers breeze through, while capable but quiet people get unfairly rejected.

Scenario interviews flip that. You don't ask candidates to praise themselves, you put them in the real work situation and watch what they do. A good salesperson shows up in how they handle a customer who is saying no, not in the line “I have a high sense of responsibility”. With the same thinking, we also use machines so that an AI Agent for business management measures the team's real performance instead of listening to pretty reports.

This way of judging through real work isn't something MONA invented. Harvard Business Review has long shown that scenario-based interviews predict job performance better than theory questions. Gartner notes that many HR departments are shifting to assessment based on real ability, and McKinsey shows that AI helps extend this approach to hundreds of candidates. To prevent rote preparation, a good system also shuffles questions across many variants and by industry, so two candidates applying for the same role won't get the same set. It is the same thinking we use when standardizing sales teams: measure by real performance, not by promises.

How the machine scores, so you can trust it

A legitimate worry: if a machine gives the scores, what guarantees it isn't scoring at random? The answer is that scoring must be transparent and backed by evidence.

Our interview system scores on a clear scale, split across several axes such as situation handling, problem-solving and attitude toward customers. Each score comes with a quote of exactly what the candidate said, plus a few-line summary the reviewer can read in half a minute. There is a pass/fail threshold set in advance, and borderline applications are always reviewed by a person rather than left to the machine alone.

Scoring with evidence matters because it gives you the right to overturn the machine's decision. You don't receive a bare score, you receive the reasons. That is the line between a trustworthy tool and a black box.

What the AI interview feels like for candidates

You may worry that candidates will find interviewing with a machine cold and lose goodwill toward the company. Done right, the opposite happens.

Candidates can interview any time they are free, without taking time off work to attend and without a week of back-and-forth scheduling emails. They are asked real scenarios, so they feel fairly evaluated instead of being rejected because of one meeting where the interviewer was tired. Many even say they like it because they get to show real ability.

The lighter chat that answers questions, sends reminders and keeps candidates from dropping out midway, we split into its own track, which you can read about in the recruitment chatbot article. The two complement each other: the chatbot keeps candidates engaged, and the AI interview evaluates ability.

Not just office hiring: many industries that interview in volume can use it

Any industry that has to interview many people for the same type of role benefits from this approach.

• Retail and F&B chains with 20+ branches: hiring sales and service staff constantly, with AI interviews keeping a consistent standard across locations.
• Customer care centers: need to measure phone-handling ability, exactly what the machine simulates well.
• Spa, aesthetics and clinic chains: hiring consultants and technicians for peak seasons.
• Factories and logistics: hiring supervisors and team leaders, where basic management thinking needs measuring before meeting in person.

What they share is the problem of “many interviews, the same criteria, and standards that drift between recruiters”. This is also why AI interviews usually come with the whole applying AI in the HR departmentrather than standing alone. They typically connect to recruitment management software to store results, and to the broader direction of business automation.

MONA uses it for real before selling it

We're not talking theory. Right now MONA is hiring Account staff with its own AI interview system, running on real ads at a real cost per candidate who passes the first round.

Candidates talk with an AI playing a difficult client, handle three real scenarios from the job, get scored on a threshold-based scale, then book a meeting with the leader for the final round. This whole outer round is about 80% automated, and people step in only at the deciding moment. For training candidates after they are hired, you can look at Khánh Hùng Academy. This is one piece of a recruitment AI Agent from MONA, sitting alongside the AI Agent suite that we built for ourselves and then opened up to businesses.

A frank word: we don't advise you to stitch together drag-and-drop tools to build an AI interview booth. They work when you test a few people, but in peak season, with hundreds of candidates arriving at once, they clog and fail easily, and nobody is accountable when they break. MONA builds custom software that carries load day and night, with a team behind it. If you want a more complete HR system, the Mona Software team also builds custom CRM, HRM and ERP management software, or a recruitment website as the gateway for receiving applications.

Under the hood of an automated AI interview: Google Meet, transcript and criteria-based scoring

Organizing an interview has always been a pile of small chores: messaging back and forth to fix a time, creating a meeting room, reminding the interviewer so they don't forget, and after the session sitting down to recall what the candidate said so you can write notes for the boss. Each is tiny, but together they add up to half an HR workday per candidate. An automated interview system hands that whole chain to the machine.

The flow works like this. A candidate who passes screening is invited to pick a slot from the interviewer's open calendar, and once chosen, both calendars sync on their own. At the scheduled time, the system creates a Google Meet room automatically: the link goes to the candidate by email, and the interviewer gets a Telegram alert just before the start and only has to click in. The interview is recorded automatically and the recording becomes a transcript, a written record of everything both sides said, with no sentence missed. The AI reads the transcript and rescores against the role's own criteria sheet, producing an objective review for the boss to approve. Whether the interviewer that day was easygoing or in a bad mood, the machine's scoring still sticks to the criteria. Two views side by side make for a fairer decision for the candidate.

The system also has one small but firm rule: a candidate who books and then doesn't show up, without notice, has their phone number and email added to a permanent blacklist. The machine keeps the schedule seriously and asks for the same seriousness back from participants.

Frequently asked questions about automated AI interviews

Do automated AI interviews replace interviewers?

Not entirely. The machine handles the outer round to screen and do preliminary scoring, while the final meeting and the decision about who to hire stay with people. This way the leader only meets candidates who have already been screened.

What if candidates cheat when interviewing with a machine?

The system guards against rote preparation by shuffling questions across many variants and asking dynamically based on the answers, so it is hard to prepare in advance. There is also still the final in-person round to cross-check.

Is AI interviewing suitable for senior roles?

The higher the role, the bigger the human part of the meeting. AI fits best in the outer round for volume-hiring roles, while for senior management it is better to let the machine do a preliminary screen and then have people interview in depth.

Is candidates' interview data safe?

A serious system stores and processes data inside the business's own system, masks personal information as required by regulations, and doesn't send it outside. Ask any provider about this point in detail.

What you can try this week

Pick the one position you interview for most, then count the hours your leader spends on the first round each month. Multiply that by their hourly pay and you'll immediately see the amount automated AI interviews can give back.

Once you have the number, we invite you to book a call with MONA. We'll show you an AI interview running for real and work out how to apply it to the exact position where you're short of people. If you want the big picture first, start with the article What is AI recruitment.

To see how this AI interview round fits into the complete package, visit the page the AI recruitment software Gấu Tuyển Dụng.