How to get hired at Mercor: AI interview and CV guide

Why Mercor's hiring process is different

Mercor does not run traditional phone screens or panel interviews. The entire screening process is handled by an AI interviewer, a video-based system that asks role-specific questions, transcribes your responses, and evaluates your performance against other candidates. You complete one interview, and that result is reused across every matching role you apply for on the platform. This means a single 20-minute conversation can open or close dozens of opportunities, so it is worth preparing for properly.

There are two distinct tracks on Mercor. The engineering and contractor track is for software professionals seeking remote contracts with US companies, typically paying $50 to $150 per hour. The AI training and data-labelling track covers annotation, writing, evaluation, and translation work, with rates starting around $12 to $25 per hour and no engineering experience required. Knowing which track you are on changes how you prepare. If you are applying for an annotation role, drilling algorithm questions is a waste of time.

How the AI interview works

The interview lasts around 15 to 20 minutes. You access it from your Mercor dashboard under the Application tab. Before it begins, you will verify your identity with a government-issued ID, and your camera and microphone must remain on throughout. The AI asks behavioural and role-specific questions, generates a transcript of your answers, and scores you across several dimensions.

The system evaluates specificity, ownership, and communication compression. Vague answers score low. Saying "I worked on a project that improved performance" is not enough, but saying "I reduced API latency by 40% by switching from REST polling to WebSockets on a team of four engineers" scores significantly higher. The AI also prefers candidates who say "I" rather than "we". Many people are trained to say "our team accomplished X", but on Mercor's system this reads as deflection. Own the specific contributions you made, even within a team context.

Mercor offers a free Practice Interview on their platform that does not affect your real results. You should take this before your official interview to get comfortable with the format.

Preparing STAR answers

Behavioural questions on Mercor typically follow the pattern: "Tell me about a time you handled conflict, led a project, dealt with a technical challenge, or made a mistake." The STAR framework (Situation, Task, Action, Result) is the best way to structure your responses.

For the Situation, keep it to one sentence. Compressed context scores higher than lengthy backstory. For the Task, explain what you specifically needed to accomplish. For the Action, this should be roughly 60% of your answer, three to five specific actions in sequence, all using "I". For the Result, quantify it. Time saved, revenue generated, users affected, percentage improvement. If you do not have exact numbers, use comparative framing such as "twice as fast" or "reduced by half".

Prepare four to six STAR answers before the interview. Practice each one under 90 seconds. Record yourself and listen back. If you would not want to hear a candidate give that answer, rewrite it.

Technical setup that affects your score

A poor recording environment can hurt your score indirectly because hard-to-parse audio makes it harder for the system to evaluate your communication. Use a wired ethernet connection over WiFi where possible, as connection drops can corrupt your recording mid-answer. Use an external microphone or headset, since built-in laptop mics create echo and background noise. Ensure bright front-facing light, because dim or backlit video makes facial tracking less reliable. Choose a quiet, distraction-free room with a clean background.

The system records your full screen for the duration, so do not have notes tabs open. Your camera must stay on, and eyes consistently leaving the frame can get flagged. Have pen and paper and a calculator available if relevant to your role.

Using retakes strategically

Mercor allows up to 3 retakes of the AI interview. Most candidates do not use this deliberately. A smarter approach is to treat your first attempt as a baseline. You will learn which question types the system asks. After the first attempt, identify which answers were vague, which ran long, and which lacked ownership language. Write tighter versions of each answer. Use your second attempt with revised answers, as this is typically where candidates see the largest improvement. Reserve your third attempt for when you have verified your technical setup and know your answers are solid.

Do not burn retakes when you are rushed or stressed. The interview will wait. Treat each attempt as a calibration round.

Preparing your CV with AI experience

Mercor's system uses AI to parse your resume for technical skills, and it prioritises your first 5 skills when ranking your profile. Everything after that matters less, so put your strongest, most relevant skills at the top. Include both your professional history and any specific AI project experience. If you have worked on other AI training platforms such as Outlier or DataAnnotation, you can include that experience, but be honest about it. List yourself as a freelance AI trainer rather than naming a specific client as your employer, and do not inflate job titles.

When describing AI training work, focus on outcomes. Mention the types of tasks you handled (annotation, evaluation, writing, coding), the domains you worked in, and any quality metrics you achieved. If you completed training or certification in machine learning, natural language processing, or prompt engineering, include it. The system looks for demonstrable expertise, so be specific about what you know and what you have done.

Preparing your CV without AI experience

If you have no direct AI training experience, do not try to fake it. Mercor's AI interview will probe your stated skills, and if you mention ML terminology you cannot explain when asked, it flags profile inconsistency. Instead, emphasise transferable skills. If you are a software engineer, highlight your coding projects, problem-solving approach, and technical communication. If you are a writer, highlight your editing, research, and language skills. If you are a domain expert (law, medicine, finance, education), highlight your professional credentials and your ability to evaluate information critically.

For the AI training track, many roles require no engineering background at all. Annotation, writing, and evaluation roles value attention to detail, clear written English, and reliability. Your CV should demonstrate these qualities through concrete examples from any professional context. Frame your experience in terms of accuracy, throughput, and quality, because these are the metrics that matter in AI training work.

Common mistakes to avoid

The most common mistakes that get candidates filtered out are using "we" language without "I", giving answers over three minutes long, providing no quantified results, looking off-camera when reading notes, restating the question before answering, and using jargon without substance. If you are on the AI training track and mention ML terminology you cannot explain when probed, it flags profile inconsistency.

One final tip from candidates who have passed: look into your camera when you are doing your interview with the AI, as if it were a human. You need to make eye contact. Treat it like a real conversation, because the system is designed to evaluate how you communicate, not just what you say.