Common questions, honest answers

A few small things go a long way:

• Turn on Do Not Disturb before you start. It keeps notifications, calls, and message previews off your screen while you're sharing it with the recruiter.
• Close apps you wouldn't want visible: personal email, social feeds, chat apps, anything you'd rather keep private.
• Pause screen share any time you need to. If something comes up mid-session, just pause the screen sharing and turn it back on when you're ready to continue. There's no penalty for it.

For coding assessments Treat the session like a normal working session, use your favorite editor, tools, and setup.

You'll receive a link by email. When you open it, you'll be asked to confirm your details and start a timed session. Depending on the role you're applying for, you'll either:

• Work in a real codebase: you get a GitHub repository (cloned from a template), review the code, fix issues, and open a pull request. You use your own editor, tools, and workflow.
• Complete a communication assessment: you'll read a passage aloud and take notes on a short audio clip, simulating the kind of async work that happens in distributed teams.

Assessments are scoped to be respectful of your time, typically between 5 and 55 minutes depending on the role.

Not really. By default, the system asks you to join with your webcam briefly, mainly so you can say hello and confirm you're a real human starting the session. After that, you can turn your webcam off at any point, or cover it from the start.

Nobody is sitting there watching a live feed of you. The signal hiring managers care about is your reasoning. Talk through the problem, and explain your decisions.

Think of it like walking a teammate through your work on a quick call. You'd explain your thinking, not perform for a camera.

Yes. Your screen and microphone audio are recorded for the duration of the assessment. This is how the hiring team reviews your work: they watch the recording and read the auto-generated transcript to understand how you approached the problem.

The recording is what replaces multiple rounds of interviews. Instead of asking you to come back and explain your code, the hiring team can simply watch and listen at their own pace.

Nothing else is captured: no keylogging, no clipboard monitoring, you can use any tools you like.

Yes, absolutely. AI tools are not banned. In fact, using them is welcome. They're part of how real engineering work happens today.

The signal we're looking for is not "can you code without AI?" It's: do you understand what the tools are doing, and can you reason critically about the output? Copying an AI answer verbatim and submitting it without understanding it is somethign anyone can do. Explaining your reasoning as you use tools is what stands out.

The same logic applies to Google, Stack Overflow, documentation, and any other resource you'd normally use on the job.

If you do use AI: any free tool is more than enough for these problems. There is no advantage to using a more expensive or powerful model. The assessment isn't about which AI you have access to, it's about how you think through the problem and reason about tradeoffs.

You're encouraged to narrate your thinking. This is the main signal the hiring team uses to understand your reasoning. But there's no rigid script or requirement to speak continuously.

Think of it as a brief think-aloud, like explaining your work to a colleague: "I'm looking at this function because it seems like the bug is in the edge case handling… let me check the tests first."

It might feel a little awkward at first, which is completely normal. Most candidates find their rhythm after the first few minutes. Some exercises include a short warmup at the start, specifically to help you get comfortable.

Assessments are scoped deliberately to respect your time. Most run between 20 and 55 minutes, depending on the role and seniority level. Some communication-focused assessments are as short as 5–15 minutes.

The assessment page will show you the expected duration before you begin, so you can plan accordingly. You can start the assessment at any time.

For coding assessments: yes. If you close the tab or lose your connection, you can reopen the same link and resume where you left off. The resume function is available for disconnections of up to five minutes. Your recording and progress are saved.

For technical communication assessments: resuming is not supported, because the structure depends on sequential steps. These assessments are usually quite short. If something goes wrong, reach out to the recruiter who sent you the link and they can issue a new one.

Technical issues happen. If your browser crashes, your internet drops, or something unexpected occurs, contact the recruiter who sent you the assessment link. They can see the status of your session and, if needed, issue a fresh link.

For coding assessments, your work in the GitHub repository is preserved regardless of recording status, so even if the recording is affected, your code is still there.

No. MarioHR explicitly tries not to penalize non-native English accents or transcription artifacts caused by accent. The assessment is designed to help evaluate technical reasoning and communication clarity, not whether you sound like a native speaker.

Hiring managers look at things like: did you identify the right problem, did you explain your reasoning clearly enough to follow, did you cover the key points. How you pronounce things is not part of the rubric.

Only authorised members of the hiring team at the company you applied to, typically the recruiter and the hiring manager. MarioHR integrates with the company's ATS (their hiring software) and makes it easy to access the output of the assessment to authorized users.

MarioHR staff do not routinely review individual recordings, unless the company opts for expert human review of the assessment output before it is sent to the hiring team.

MarioHR assessments are built specifically to be reviewed, not just logged. The combination of a recorded walkthrough and objective output means the hiring team gets meaningful signal without you having to write a lengthy project from scratch.

You may request a live interview or other alternative format directly from the hiring company in accordance with their own policies and applicable law, this is a decision made by the employer, not MarioHR. Contact the recruiter or hiring team that sent you the assessment to make any such request, including requests for additional time, different tools, or other accommodations. Making an accommodation request is not treated negatively.

Most teams have a consistent problem: by the time a candidate reaches an engineer interview, nobody really knows whether the conversation will be worth it. MarioHR gives the hiring team actual signal before that point, so the later conversations are more focused and better informed.

In practice, this means:

• Consistent evaluation criteria: every candidate gets the same chance to demonstrate their skills, it's not based on whoever happened to be on-call for interviews that week.
• Context you can actually act on: a recording of the candidate walking through their thinking, plus a structured output, is far more useful than a recruiter summary of a phone call.
• A better candidate experience: candidates get a realistic task with their own tools, a clear time expectation, and a fair evaluation process. They can take the assessment at any time, and may get a chance to receive feedback on their work before the next round.

No, and it's not designed to. Engineers still review the recording and the code. They still run later-stage interviews. MarioHR just means those conversations happen with candidates who've already demonstrated they can think through a real problem.

The final call is always yours. MarioHR produces a structured output and a recording; your team decides what to do with it.

The reporting process works in layers:

• Summary: After a candidate submits, the transcript and/or codebase, the data is sent to our summary engine via AWS Bedrock. It interprets the submission against a detailed list of items covering code quality, problem diagnosis, thought process, technical communication, and awareness of tradeoffs. This is meant to be a summary of the candidate's work, absolutely not a final verdict.
• Human review: Your team then reviews the recording, read the transcript, and look at the pull request to make a decision.
• Expert human review: Optional expert human review is also available. Our team can review the assessment output and provide a detailed report to the hiring team.

MarioHR is built to live inside your existing workflow, not sit next to it. Current native integrations:

• Ashby: Move a candidate to the assessment stage and MarioHR automatically sends the invite link. The output is sent to the candidate's Ashby profile.

A full REST API is also available if you have a custom ATS or want to trigger assessments programmatically. See the Ashby integration guide as an example.

MarioHR offers two assessment types that together cover a wide range of roles:

• Coding assessments: for any technical role where code review, debugging, or code quality is a meaningful signal: backend engineers, full-stack, DevOps, platform, and similar.
• Technical communication assessments: for roles where clear writing and listening skills matter.

Custom exercises are available if your role has specific requirements not covered by the standard catalog. Reach out to dev@mariohr.com to discuss your needs.

Candidate experience is a design principle in MarioHR. A few specific choices that reflect this:

• Candidates use their own tools, IDE, and AI assistants, no artificial lockdowns that would never exist on the job.
• Assessments are scoped to be short. Most run under an hour. The time commitment is communicated upfront.
• There's accessibility support built in: light/dark mode, font size and family controls, keyboard navigation.
• Webcam is optional after the initial check-in.
• The hiring process is managed by the company, you choose which steps to include before and after the assessment. The candidate experience starts from marketing, a well crafted job description, and ends with the hiring decision. Most of the time, the assessment is only a small, but key step in the process.

Yes. For coding assessments, candidates work in a dedicated GitHub repository. The pull request they open is linked directly in the report.

You can also add external reviewers (e.g. a senior engineer on your team) as GitHub collaborators directly from the report page, so they can leave code review comments without needing to access the MarioHR dashboard.

Yes. Each assessment has a shareable report link. Access can be controlled in two ways:

• Password-protected link: anyone with the link and password can view the report. Good for sharing with hiring managers or interviewers who aren't in MarioHR. Passwords are report-specific and can be rotated or disabled by a recruiter admin at any time. Password access is disabled by default.
• SSO-only: access is restricted to people who sign in with your company's Google domain and each account has to be approved individually.

All files are encrypted both at rest and in transit. Session recordings are deleted automatically on a rolling schedule by lifecycle policies. The raw recording is deleted within 30 days. Processed files (video, transcript) are permanently deleted by 90 days. No manual step is required. If you'd like to get your session deleted sooner, please contact us at dev@mariohr.com.

Recordings are encrypted at rest on AWS S3. Access is controlled via signed URLs with keys. Recordings are never publicly accessible. All data in transit uses TLS.

MarioHR is actively working toward SOC 2 and ISO 27001 certification, so that third-party auditors can independently attest to these practices. We take these certifications seriously and have built toward them from the start.

Yes. Candidates provide explicit consent before starting any assessment. Recordings are deleted automatically after 90 days. Europe-based storage is available for european clients. Candidates can request earlier deletion of their data by contacting dev@mariohr.com.

The Privacy Policy is linked at the point of consent and available on the website.

No. Assessment recordings, transcripts, and candidate data are not used to train AI models.

No. MarioHR does not perform facial recognition, emotion detection, eye tracking, keystroke dynamics analysis, or any other form of biometric or behavioral profiling.

MarioHR supports SSO (Single Sign-On) via Google today, with role-based access control. A full REST API is available for custom integrations.

For specific security requirements or compliance questions, reach out at dev@mariohr.com.