What Recrutly’s Arabic Interviews Do—and What Remains Unproven

Recrutly advertises Arabic support for conducting interviews, not just an Arabic interface. Onrec’s product coverage describes a “full Arabic-language experience,” including interview scheduling and interviewing. That does not establish accurate transcription or comparable scoring across Arabic dialects. Candidates should confirm Arabic is enabled for their particular interview; employers should test the language workflow and resulting assessments before using AI scores to filter applicants.

Select the evidence your team has checked; the result identifies the next deployment decision.

Recrutly Runs Employer-Configured, Live AI Interviews

Recrutly is an employer-side hiring platform, not simply a mock-interview coach. Its documented workflow connects job setup, resume screening, interviews and candidate comparison. Recruiters define role requirements and evaluation criteria, review AI-generated questions, and decide which candidates move forward.

According to its AI interview page, a candidate:

1. Receives a link and completes the application form.
2. Selects an interview time slot.
3. Joins a live voice-and-video interview with a Recrutly-hosted AI interviewer.
4. Answers role-specific questions and adaptive follow-ups based on their responses.

The vendor says interviews last up to 30 minutes, are recorded and transcribed, and run with proctoring throughout. This differs from a fixed one-way interview in which every applicant records answers to the same prompts without follow-up questions.

Adaptive follow-ups make language handling relevant at more than the final transcription stage. If an Arabic answer is misunderstood, the next question could pursue the wrong subject or fail to explore relevant experience. A pilot should therefore inspect the conversation itself, not only whether the final transcript looks readable.

The employer’s configuration also matters. A platform-level statement about Arabic does not confirm the language selected for a particular vacancy, the questions the recruiter approved, or the instructions sent to its applicants. Candidates need an answer from the inviting employer rather than an assumption based on the vendor’s marketing.

The candidate report documentation lists an overall AI score, section scores, a separate communication score, transcript, audio/video recording references, proctoring signals, strengths, risks and a recommendation. It describes evaluation across skills, communication, experience, role fit and employer-defined criteria.

Those descriptions do not explain precisely how the communication score is calculated. Do not assume it measures only answer content—or that it measures facial expressions or emotions. Ask which inputs actually affect the score and how the employer intends to use it.

Arabic Availability Is Not Dialect Validation

Onrec’s coverage reports the advertised language capability; it does not demonstrate its accuracy. The product and fairness pages cited here do not provide dialect-by-dialect accuracy results or an Arabic-specific scoring validation study. No Arabic transcription accuracy figure or dialect-level scoring benchmark is available in those cited materials.

That leaves separate questions about access, interpretation and evaluation. Arabic instructions may work while a transcript loses meaning. A transcript may preserve the answer while an assessment assigns an unexplained communication penalty. Neither an Arabic interface nor a fluent-sounding interviewer resolves those questions.

Employers should request evidence matching their intended applicant pool and use case. Results for a language variant outside that pool would not establish performance for the dialects candidates actually speak. Likewise, a demonstration of transcription would not establish that scores reflect job-related evidence.

Check	What To Request Or Test
Arabic workflow	Check invitations, scheduling, instructions, questions and follow-ups. Separately confirm transcript and report languages.
Relevant dialects	Use consenting pilot participants from the intended applicant pool, alongside Modern Standard Arabic where appropriate.
Mixed-language answers	Include English job titles, product names and technical terms within realistic Arabic answers.
Score comparability	Compare equivalent answer content across language variants against a job-related human rubric.
Communication and proctoring	Obtain scoring inputs and flagging rules, including how disputes and technical failures are reviewed.

These checks should not become a requirement that every applicant speak identically. Their purpose is to find out whether the assessment preserves relevant evidence when candidates express it differently. If a role genuinely requires a particular language skill, the employer should define that requirement separately from an unexplained platform score.

A Pilot Should Trace Answers Through To The Report

Start with the interview configuration the employer actually plans to use. Confirm the approved questions, expected language, evaluation criteria and report settings before asking participants to complete it. Testing a different configuration would leave uncertainty about the experience applicants will receive.

Use consenting participants whose speech reflects the relevant applicant pool. Have Arabic-speaking reviewers assess whether the transcript preserves meaning, not just whether its spelling is tidy. Review the recording alongside the transcript when wording is disputed. A polished transcript can still omit a qualification, reverse a statement or misrepresent a technical term.

For example, an Arabic-speaking engineer might explain a project in Arabic while using English software terminology. Reviewers should check whether the report preserves the technical evidence and whether follow-up questions reflect what the person actually said. The assessment should not be accepted merely because the conversation continued without an obvious interruption.

Record where a problem entered the workflow. Was the spoken answer misunderstood? Did the transcript capture it correctly but the report summarize it incorrectly? Did a relevant answer reach the report but receive an unexplained low score? Those are different failures and require different explanations from the vendor.

For score comparisons, use equivalent substantive answers and a job-related human rubric. Keep the expected evidence consistent while examining language variants. If content, experience and language all change at once, the comparison cannot isolate whether language handling contributed to the difference.

Investigate unexplained differences before setting cutoffs. Ask for the evidence behind each affected assessment, including which answer supported a stated weakness or recommendation. Where the report cannot be reconciled with the recording and rubric, do not treat the score as established evidence of applicant quality.

A small pilot can expose errors; it cannot, on its own, establish fair scoring across the applicant population. The cited materials do not supply a validated sample size or an acceptable error threshold for Arabic interviews. Employers should not invent either from a successful demonstration or a handful of satisfactory reports.

Keep the deployment decision narrower than the pilot result. Confirming that Arabic questions and instructions work supports a conclusion about workflow availability. Confirming selected transcripts supports a conclusion about those tested interviews. Neither establishes reliable automated rejection across all Arabic-speaking applicants.

Communication Scores And Proctoring Need Separate Review

The report’s separate communication score deserves its own explanation. Request the inputs it uses, the employer’s selected criteria and the evidence shown to reviewers. A candidate’s technical answer, speaking style and ability to perform a language-dependent job task should not be treated as interchangeable without a defined assessment rationale.

Ask how mixed Arabic and English answers are handled within that score. Also ask what happens when a candidate repeats an answer, corrects a misunderstood term or asks for clarification. These are test cases to examine, not documented Recrutly scoring rules; the cited pages do not establish their effects.

Proctoring signals are another distinct output. The vendor says proctoring runs throughout the interview, and the report documentation lists proctoring signals. Those statements do not, by themselves, explain every monitored event or what any particular flag proves. Obtain the rules and agree how reviewers will handle disputed flags, speech barriers and technical failures.

Recrutly’s fairness policy says recruiters can accept, modify or disregard AI insights and retain final hiring authority. Make that operational: require reviewers to inspect disputed transcripts and flags rather than simply approve a recommendation.

Human final approval is not a substitute for validating the assessment. A reviewer who sees only a score or summary may miss the underlying language error. Give the responsible reviewer access to the relevant evidence and a way to correct the hiring record. Our AI hiring vendor validation checklist covers the broader checks.

Candidates Should Confirm Their Interview’s Language Settings

If you have received an invitation, ask the employer whether Arabic is enabled for that specific interview and whether there are restrictions on dialect or mixed-language answers. The distinction matters when your usual explanation includes English product names, technical vocabulary or job titles.

Ask what is scored, particularly under “communication,” and what proctoring monitors. The product descriptions establish that these outputs exist, but they do not provide a complete explanation of their calculation or monitoring rules. The employer should explain the assessment it has chosen to use.

Before starting, identify who handles transcription errors or technical problems and whether another interview format is available if needed. Also ask who can access the recording and report, and how long each is retained. Getting a named contact is more useful than discovering after the interview that nobody has been assigned to review a disputed answer.

Prepare concrete examples of your skills and results, and practise explaining them aloud in the intended interview language. Because the vendor describes adaptive follow-ups, prepare to explain your own contribution, reasoning and outcome rather than memorize a single uninterrupted speech.

Do not change dialect solely to chase an unknown score. There is no dialect-by-dialect scoring evidence in the cited materials that would justify recommending one as a scoring advantage. Follow the employer’s confirmed language requirements and use terms that accurately describe your experience.

If a language or technical problem occurs, report it promptly. Describe the affected question or answer and ask the recruiter to review the relevant recording or transcript. For a mistranscribed technical term, explain what you said and why the correction changes the evidence about your skills. You do not need to guess how the scoring system works to identify an inaccurate record.

For barriers related to disability, see our AI interview accommodation guide. An alternative format should be discussed with the employer; the cited Recrutly workflow does not establish that every applicant will automatically receive one.

The Employer Controls Key Retention Decisions

Recrutly’s Privacy Policy says candidate-data retention follows the controlling customer organization’s instructions and settings. Its privacy guidance also lists plan-specific limits on evaluation storage after a plan ends and says deletion requests are coordinated with the controlling organization.

These are not a single retention period that applies to every employer or every interview record. Ask the employer for its own notice and the retention periods for recordings, transcripts, evaluations and related hiring records. Do not assume a limit described for evaluation storage after a plan ends applies identically to every record type.

Candidates should direct questions about their particular interview to the employer controlling the process. HR teams should ensure their candidate notice matches the settings and practices they actually use, including who can review reports and how a correction or deletion request is routed.

Applicable notice, privacy and accommodation requirements depend on jurisdiction. Arabic language support does not establish compliance with the rules governing an employer or its applicants. This is a product explainer, not an endorsement, an independent performance test or legal advice.