Privacy Policy

What this policy covers

This policy explains how Interview Command handles information when you browse the site, create or use an account, use social sign-in, use Resume Lab, save practice progress, use speech-practice features, submit an outcome survey, or purchase a paid plan.

Information you choose to provide

Email. If you submit an email for access, account creation, updates, checkout, or purchase recovery, the site may store or process it as needed to provide access and administer the product.

Social sign-in. Available social sign-in options are shown on the account page. When you choose one of those providers, it may return a provider-specific account identifier and, depending on provider permissions, an email address, display name, profile image, and email-verification signal. Interview Command uses those values to create or locate your account and restore eligible access. The service does not intentionally retain provider access or refresh tokens after sign-in.

Résumé. Resume Lab sends the file to this site's server over HTTPS and processes it temporarily in application memory to extract interview-preparation facts. The current Resume Lab does not intentionally write the raw résumé to application storage and does not send it to an external AI/LLM API.

Practice data. Progress, onboarding choices, résumé plans you choose to save, and speech-attempt history may be stored in your browser using local storage or cookies. For a signed-in user with an active paid plan, Account Sync is optional and off by default. If you explicitly enable it, Interview Command can store question progress/attempt counters, delivery metrics such as WPM/hesitation/pause/route-coverage data, and Strict Simulation summary metrics in PostgreSQL so that telemetry can be merged across signed-in devices. Account Sync does not currently include Resume Lab facts/plans, VIP target-workspace snapshots, transcripts, raw audio, or raw video.

Role-library and acquisition analytics

First-party site measurement records the website hostname, page path without query strings, a coarse screen-size category, and capped JavaScript-error counts without error messages or candidate content. Browser privacy opt-outs are respected. Owner pages are not automatically measured. Marked QA sessions and events recognized as owner activity are excluded from the app-hosted default growth report.

To understand which preparation paths candidates use and which role packs or content to improve, Interview Command may record limited first-party events such as landing-page views, free-practice starts, Start Free actions, free-question completion milestones, role-library filter selections, researched-pack opens, readiness-check completion, signup completion, checkout intent, verified checkout starts, purchases, and normalized role-library search terms. Acquisition records may include campaign tags supplied in the link (such as source and content ID) and a random first-party browser-session identifier used to de-duplicate funnel steps. The acquisition session identifier is not a transcript, résumé, raw audio, or raw video and is not used to make hiring decisions. Search telemetry is delayed and de-duplicated; obvious email addresses, URLs, and phone-like number strings are removed before storage. A zero-result role search is counted as unsupported profession demand only when the catalog is unfiltered. This telemetry is used for aggregate product and acquisition analysis, not to determine employment eligibility or make hiring decisions.

Human Review submissions

If you use the VIP Priority Pass Human Review benefit, Interview Command stores the practice response or transcript you submit, the review title and target role you provide, optional reviewer-focus notes, limited session-summary data you choose to include, the submission and due timestamps, and the human reviewer’s written feedback and exact pointers. An authorized human reviewer can access that material for the purpose of providing the requested coaching.

Human Review content is not used as an employer assessment and is not shared with an employer through this feature.

Affiliate referral attribution

If you arrive through an activated affiliate link, Interview Command may store the approved affiliate identifier in a short-lived first-party HttpOnly browser cookie for up to 30 days so that identifier can be forwarded to Lemon Squeezy checkout. The cookie is used for referral attribution and the related promotional discount; it is not used to infer hiring preferences or for cross-site advertising profiling by Interview Command.

Lemon Squeezy may separately process affiliate and referral information under its own privacy practices when checkout or the Affiliate Hub is used.

Optional AI answer review

When you choose “Review this answer,” Interview Command sends the answer text, the previous answer to that question when available, and the question’s reference material to OpenAI for provisional coaching. Audio, your account email, and unrelated practice records are not included in this request. Saving an answer alone does not trigger this optional review. Avoid including confidential employer or personal information.

The request disables API response storage. This does not promise zero provider retention; OpenAI’s applicable API data policies still apply. Interview Command processes these answer texts in memory for the review and does not intentionally log them. The returned feedback and review history are saved with your browser’s practice record. A server-side copy of the feedback (which can include answer excerpts), an account identifier hash, request hashes, and usage totals are also retained to reuse existing reviews and enforce review allowances. Full submitted answer texts are not stored in the budget ledger. AI feedback can be wrong and does not establish interview readiness.

Payments and order records

Lemon Squeezy acts as merchant of record and handles payment credentials, payment-method collection, applicable indirect taxes, receipts, refunds, chargebacks, and related checkout processing under its own terms and privacy practices. Interview Command does not receive or store your full card details.

The current Free tier does not require a payment method. Current paid offers are fixed-duration one-time passes. Interview Command receives only the limited order information needed to verify payment, grant access, enforce the selected pathway scope and access duration, support refunds, and restore eligible purchases.

Lemon Squeezy webhooks may provide limited order data such as email, order identifiers, product/variant, payment or refund status, timestamps, and access-linking metadata.

Paid access recovery and account linking

Returning customers can restore paid access with the purchase email and the numeric Order # shown on the Lemon Squeezy receipt. Internally, Interview Command maps that visible order number to Lemon Squeezy's order identifier after a verified webhook is received. If you are signed in, a verified active purchase can also be linked to your Interview Command account so eligible access can be restored after account login.

Account sessions

Social-account login uses an opaque browser session cookie. The server stores a cryptographic hash of that account-session token together with the account identifier, timestamps, and expiration. Provider identity records retain the provider name and provider-specific subject identifier and may retain the provider-returned email and verification status needed for safe account linking.

Outcome survey

The optional post-interview survey does not require your email. Your browser creates a random identifier; the server stores only a SHA-256 hash of that identifier with your self-reported outcome, mock-session count, locally detected recorded-attempt count, whether Premium was active on that browser, and optional comments. These responses may later be analyzed in aggregate to understand product efficacy. Individual survey responses are not presented as verified hiring records.

Audio, camera, and Strict Simulation

For ordinary browser-only practice, speech analytics can use a browser-generated transcript, local timing/pause signals, and derived acoustic measurements such as estimated fundamental-frequency range, within-answer loudness variation, and microphone clipping. These acoustic measurements are experimental descriptive evidence: they are not calibrated as a confidence, emotion, accent, personality, employability, or hiring score and do not currently determine readiness. For eligible signed-in premium spoken attempts, the current recording, available browser transcript, and measured delivery/acoustic signals may be sent to InterviewCommand over the same origin for transient in-memory practice evaluation. That evaluation endpoint is designed not to write raw audio or raw transcripts to the application database. Derived metrics and structured coaching/evidence from the attempt may remain in browser-stored practice history. Browser speech-recognition features may also be processed locally or by your browser/vendor depending on the device and browser. Raw Strict Simulation camera video is not intentionally uploaded to the InterviewCommand server.

Optional Zara Voice calibration contribution

After a completed eligible Zara Voice Simulation, InterviewCommand may offer an optional calibration contribution. Nothing is uploaded to the calibration corpus unless you affirmatively check the consent box and submit. If you opt in, the contribution can include your primary-answer voice recordings, browser transcript, server-verified question context, timing and delivery measurements, and derived acoustic measurements. A voice recording can identify a person. We therefore describe these records as pseudonymous, not anonymous or fully deidentified.

The calibration tables are designed not to store your InterviewCommand account ID, email address, purchase/order ID, or simulation run ID. Authorized calibration raters may listen to the recording and review the transcript, question, and non-acoustic measured delivery evidence. During independent rating, analyzer predictions, derived acoustic telemetry, and other raters' labels are withheld from the rater. Calibration records are kept separate from hiring/outcome survey data and are not used as employer assessments or to infer confidence, emotion, personality, intelligence, accent quality, employability, or hiring likelihood.

Declining calibration has no effect on your account, access, scores, receipts, purchased services, or training recommendations. When you contribute, a random batch ID and private withdrawal code are saved in your local receipt; the server stores only a SHA-256 hash of that code. Presenting the matching batch ID and code through the supported withdrawal flow deletes the server-held contributed batch and its recordings, transcripts, ratings, and adjudications. A previously downloaded authorized corpus export cannot be automatically recalled by that later server-side deletion.

When you deliberately start Strict Simulation, the browser requests camera and microphone permission and may analyze camera frames locally for face presence, multiple-face detection, head-position, and visual-attention coaching signals. The current visual-analysis implementation uses MediaPipe Face Landmarker loaded from third-party library/model hosting. MediaPipe's package documentation states that usage or performance metrics may be sent to Google. Interview Command stores only a limited simulation summary in browser local storage, such as completed answers, environment departures, response timing, and sustained attention-event counts; it does not intentionally store raw camera frames on the Interview Command server.

Visual-attention, focus, fullscreen, and face-presence signals are coaching aids only. They are not biometric identity verification, a cheating determination, an employer score, or a hiring prediction.

Technical and security data

Hosting infrastructure may process ordinary request metadata such as IP address, timestamp, user agent, and security/rate-limit information. Interview Command does not intentionally place résumé contents into application logs.

Retention

The raw Resume Lab file is intended to exist only for temporary processing. Browser-stored practice information remains on your device until you clear it. An approved affiliate-attribution cookie is configured for up to 30 days. Short-lived email-verification records expire after use or timeout and may be cleaned up automatically. Server-side account, social-identity, session, opt-in account-synced training telemetry, lead, access, order-entitlement, Human Review submission/feedback, outcome-survey, security, and refund-status records may be retained as needed to operate the service, prevent abuse, deliver purchased services, honor access or revocation, and maintain business records. Opt-in calibration recordings and their pseudonymous answer records may be retained for the calibration/validation program unless the contribution is withdrawn; raw calibration audio is excluded from the finalized corpus export used for analyzer evaluation.

Sharing

Interview Command does not sell résumé contents or personal information. Information may be processed by providers required to operate the service, including hosting infrastructure, Lemon Squeezy, the transactional email provider used for passwordless verification, and the social sign-in provider you deliberately choose for authentication. If you explicitly contribute a Zara Voice calibration sample, authorized human calibration raters may access the contributed audio, transcript, question context, and measured delivery evidence solely for calibration/rating work.

Your choices

You can use the sign-in methods currently offered, choose not to upload a résumé, remove sensitive details before upload, edit extracted résumé facts, decline camera or microphone permission and skip Strict Simulation, clear browser-stored progress and simulation history, keep Account Sync off, turn it off after enabling it, delete the synced cloud copy without deleting local browser data, decline the outcome survey, decline the optional Zara Voice calibration contribution without affecting your training, and use the private withdrawal code from a contributed receipt to request deletion of that calibration batch.

Support

Public support information is available at /support. Support email: suppor@interviewcommand.com.

Adam coaching and supervised tester assignments

When you opt into Adam voice practice, microphone audio is sent to OpenAI for conversation and transcription. Adam uses an AI-generated voice. Account-linked transcripts, exercises, coaching observations and saved checkpoints expire after 30 days; your optional Adam Profile remains until you remove it. Coaching observations are provisional and are not verified credentials, hiring assessments or proof of mastery.

Supervised tester assignments require separate consent to entire-screen recording without system audio. Screen segments, tab/fullscreen condition events, application information and submitted feedback are stored for 30 days for human review. Close personal or confidential material before sharing. Stopping sharing pauses the assignment. Browser condition signals do not establish cheating. Authorized reviewers can access and export evidence; a downloaded export cannot be recalled automatically. Pilot calibration and release-review imports also expire after 30 days.

Payment records and pseudonymous cohort budget commitments are kept separately from screen recordings to prevent duplicate payments or exceeding the approved compensation budget. Usage and allowance journals can outlive deleted conversations for cost controls and financial accountability. No payment is automatically sent, and candidate information is not automatically used to train model weights. Pause Adam and use “Manage this practice” to delete the selected saved conversation, exercises, coaching observations and receipt. The dashboard also removes that receipt from the current browser; other browsers may still contain their local receipt indexes. Removing your Adam Profile is a separate action. Account deletion and other data requests are handled through support; there is currently no automatic delete-account control. Financial records and usage/allowance journals are retained separately. Local browser recordings, separately submitted review or calibration copies, and downloaded exports require separate deletion by their holders.