AI Recruiting RFP Question Bank

Q1: Which ATS systems does your platform integrate with? For each, specify whether the integration is certified by the ATS vendor or built on a generic API connection.

Q2: For our specific ATS, list every field your integration writes to in the ATS candidate record. Provide the exact field names as they appear in the ATS.

Q3: What data does your integration read from the ATS before the interview — and what does it do with that data?

Q4: How is the AI interview triggered? Can it fire automatically from an ATS stage change, or does it require a manual action by a recruiter?

Q5: How does your integration handle ATS tenant-specific custom fields? Who configures the field mapping?

Q6: What is your process when our ATS releases a new version? Who is responsible for updating the integration, and what is the typical lag time?

Q7: Provide the name and contact information of three customers currently running your integration with our specific ATS version.

Q8: Does your integration create, merge, or link candidate records? Describe exactly what happens to the ATS candidate record after an AI interview is completed.

Q9: Describe your scoring methodology. Is scoring rubric-based (predefined criteria per question) or algorithmic (model-derived ranking)?

Q10: Can interview scripts and scoring rubrics be configured at the job level — or is a single configuration applied across all requisitions in our account?

Q11: Can we see an example of a completed interview result as it appears in the ATS — not in your platform?

Q12: How long does it take to create, test, and deploy a new interview configuration for a new role type?

Q13: Who controls configuration changes after go-live — our team, your team, or a shared process?

Q14: How do you handle candidates who request a human alternative to the AI interview?

Q15: Can you generate per-question score breakdowns for individual candidates? Is this accessible to the candidate on request?

Q15a: Does your platform support both video interviews and phone interviews? Confirm that both are production-ready today — not on the roadmap. Does the scoring rubric apply equally to both modalities, or are there differences in what data is captured and scored depending on whether the interview is video or phone?

Q15b: Describe your fraud prevention capabilities. What mechanisms does your platform use to detect when someone other than the registered applicant is conducting the interview? Specifically: (a) For phone interviews, what behavioral or voice-based signals are used to detect proxies? (b) For video interviews, what visual analysis is applied? (c) How are fraud flags surfaced to recruiters, and what happens to the candidate record when a flag is triggered?

Q16: Provide your most recent SOC 2 Type II full audit report (not the attestation letter). What is the scope of controls included?

Q17: Do you have a third-party bias audit? If so, who conducted it, when, and on what dataset? Provide the methodology summary.

Q17a: How is your scoring rubric defined, and who controls the scoring criteria? When a candidate's response reveals a protected class characteristic — religion, pregnancy, disability, national origin — what does your platform do with that signal? Is it flagged and excluded from scoring, or does it pass through to the rubric evaluation? Provide a specific technical description, not a general policy statement.

Q17b: Does your platform have a mechanism to detect and flag responses that contain information that cannot legally be used in an employment decision? If so, how is that flag communicated to the recruiter, and what happens to the scoring for that response?

Q17c: Describe how a candidate opts out of the AI interview. Is opt-out available at any point during the interview — including mid-call — without the candidate needing to know a specific phrase, code word, or sequence? Demonstrate this in a live walkthrough.

Q18: How do you comply with the Illinois Artificial Intelligence Video Interview Act (AIVIA) and similar state laws?

Q19: Do you produce EEOC adverse impact data for AI interview completions and pass rates, broken out by demographic category?

Q20: What data residency options do you offer? Where is candidate data stored by default?

Q21: What is your candidate data deletion process and timeline after contract termination? What format is data returned to us in?

Q22: Have you been involved in any EEOC complaints, state agency investigations, or litigation involving your AI scoring system? If so, provide a summary.

Q23: Do you comply with GDPR Article 22 (automated decision-making)? What is your process for handling a data subject access request?

Q24: What is your platform completion rate for roles comparable to ours? Provide the definition of completion rate you use (denominator: invited, link-clicked, or started).

Q25: What channels do you support for interview delivery — SMS, email, phone, web? What is the completion rate by channel for our candidate profile?

Q25a: How is the AI interview scheduled with the candidate? Specifically: can a recruiter or the platform itself book the AI interview live while a candidate is on the phone — before the call ends? Or does the scheduling process involve sending a link that the candidate must open and act on after the call? Provide completion rate comparison data between link-sent invitations and interviews booked live on a call, if available.

Q25b: Do you support government-issued ID verification? If so: (a) At what point in the process is verification requested? (b) What document types are accepted? (c) Is verification required or optional? (d) What happens if a candidate declines to verify? (e) How is the verification result reflected in the ATS candidate record?

Q25c: Describe your multilingual capabilities in detail. (a) What languages are supported? (b) Does the candidate need to select their language before the interview begins — or can they switch languages mid-call? (c) If a candidate begins in English and immediately asks to continue in Spanish, what happens? (d) Does the scoring rubric apply equally across all supported languages, or are some languages limited in what scoring data is returned? (e) Is multilingual support included in the base contract or billed per language?

Q26: Provide mobile completion rate data. What percentage of interviews are completed on mobile devices?

Q27: What is your process for candidates with disabilities who cannot complete the AI interview format?

Q28: What languages does your platform support for interview delivery and scoring? Is the scoring rubric validated across all supported languages?

Q29: What does the candidate-facing interview experience look like — provide a live test account or walkthrough for our evaluation team.

Q30: Provide your complete pricing model — base license, per-interview fee, volume tiers, and any usage caps.

Q31: What implementation and onboarding fees are included vs. billed separately? Provide a detailed implementation cost estimate for our environment.

Q32: Are ATS integration setup and configuration included in the contract, or billed separately? Is there an ongoing integration maintenance fee?

Q33: What are the fees for post-go-live configuration changes — new interview scripts, rubric updates, new role types?

Q34: What is the renewal pricing model? Are price increases capped, and at what rate?

Q35: Provide a total cost of ownership estimate for year 1, year 2, and year 3, assuming our hiring volume.

Q36: What is the median elapsed time from contract signing to first production interview for your last 10 customers on our ATS?

Q37: Provide a detailed implementation timeline for our environment, including IT security review, ATS integration configuration, UAT, and recruiter training.

Q38: Who owns implementation — your professional services team, a third-party integrator, or our internal IT team?

Q39: What information and access do you require from us to begin implementation?

Q40: What is your process if implementation stalls — for example, if ATS field mapping requires more time than estimated?

Q41: What does UAT look like for your platform? Who defines pass/fail criteria?

Q42: What are your support hours? Are they aligned with our operating hours and time zones?

Q43: What is the SLA for critical (P1) issues — platform outage, integration failure, data loss? What is the escalation path?

Q44: What is your SLA for standard support requests (P2/P3)?

Q45: Who is our named account manager? What is their current book of business size?

Q46: What is your process for proactive outreach — product updates, integration changes, API version changes that may affect our setup?

Q47: Provide your 12-month and 24-month customer retention rates and the definition used to calculate them.

Q48: Who owns the candidate interview data generated on your platform? Does your privacy policy or terms of service grant you any rights to use that data for model training?

Q49: What data export options are available — format, frequency, and access mechanism?

Q50: If we terminate the contract, what happens to our data? What is the deletion timeline? What is the format for the final data export?

Q51: Are there contractual provisions that allow you to retain aggregated or de-identified data derived from our interviews after contract termination?

Q52: What is the minimum contract term? What are the early termination provisions?