Privacy Policy

Verli is operated by Inkvi LLC, 100 N Howard St Ste R, Spokane, WA 99201, United States (“Inkvi LLC”, “Verli”, “we”, “us”). Inkvi LLC is the controller of the personal data described in this policy. This Privacy Policy explains how we collect, use, and share information when you use the Verli websites, browser app, Chrome extension, desktop app, free online tools, and related services (the “Service”). By using Verli, you agree to the practices described here.

Information we collect

Account information

When you create an account we collect the identifiers you provide to our authentication provider (Clerk), such as your email address, name, and authentication credentials. We do not see or store your password directly.

Audio you capture

Verli processes microphone and system audio in real time to produce transcripts and translations. After you choose Continue, captured audio, language settings, and any preferred terms or context notes you enabled are streamed to our transcription partner (Soniox) solely to generate transcripts and translations. Verli does not store raw audio on our servers after the session ends.

When speaker voice matching is available, Verli sends one short audio clip per detected speaker to the Verli voice-classifier service hosted by Fly.io. The clip is analysed to infer a voice type, such as a higher or lower voice, so that a matching synthetic voice can be chosen for the translation. No voiceprint or speaker identifier is created or stored, the label is not linked to your account, and the clip is discarded as soon as the label is returned.

Audio and video file translator

When you use the public audio and video file translator, your browser processes the original file locally and creates a compressed audio excerpt of about three minutes. Verli uploads only that excerpt to Soniox for temporary transcription and translation processing, not the complete original file. Verli deletes the Soniox transcription and its uploaded audio before delivering a result. We retain the bilingual text result and limited job metadata in MongoDB Atlas for no more than 24 hours so you can recover the result in the same browser tab. We do not retain the source media or original filename.

Free online tools

Our website offers tools you can use without an account. In the live voice translator, your microphone audio is streamed from your browser directly to Soniox using a short-lived key we issue, and Verli does not store that audio; if you enable spoken translation, the translated text is sent to a text-to-speech provider listed below. In the YouTube transcript translator, the captions of the video link you submit are retrieved on your behalf and sent to Google Gemini for translation. The retrieval request is relayed through a third-party proxy provider, which receives the video link but no account information, and the result is cached on our servers for a short period. For both tools we record your IP address, or a one-way hash of it, to enforce usage limits and detect abuse.

Chrome extension

Clicking the Chrome extension opens the full Verli console in its own tab. Starting capture requires a Verli account. The primary action opens Verli's hosted sign-up or sign-in flow and returns to the same extension tab. After authentication, choose Microphone, Desktop audio, or both. Desktop audio opens Chrome's picker and requires a Chrome tab with Share tab audio enabled. Microphone-only capture does not open the tab picker. The extension does not collect a selected tab's title or URL.

When you start a signed-in extension session, microphone audio, selected-tab audio, or both are streamed directly to Soniox for live transcription and translation, and normal usage-session records and deducted minutes are associated with your Verli account. Microphone capture is off by default. Chrome asks for microphone permission when you select it. If you also select Desktop audio, Verli mixes microphone audio with the selected tab for the same transcription session. Verli does not retain raw audio. If you choose to save a meeting, generate a summary, or share a transcript, the account-linked storage and third-party processing described below apply. Purchase and subscription-management actions open RevenueCat or Stripe pages in a browser tab. Payment card details are not sent to or stored by the extension. The extension stores a random installation identifier and a rotating refresh credential in trusted Chrome storage. A five-minute access credential stays only in the background worker.

Verli's use of information received from Google APIs will adhere to the Chrome Web Store User Data Policy, including the Limited Use requirements.

Transcripts and share sessions

If you create a share session, the transcript text, language settings, and session metadata are stored temporarily so viewers can follow along. Share sessions and their transcript segments are deleted automatically 24 hours after the session is created. If you generate an AI summary, transcript text and optional context notes are sent to Google Gemini. If you use Read aloud, translated transcript text is sent to the selected text-to-speech provider, which may be Google Gemini, xAI Grok, Lemonfox, or Google Cloud Text-to-Speech.

Saving meeting transcripts is enabled by default unless you previously turned it off. When it is enabled, completed transcript text, language settings, timestamps, and speaker labels are stored in your Verli account so you can view them in History. Raw audio is not saved with the meeting. This setting applies only to new recordings and can be turned off at any time.

Knowledge-base presets and settings

If you save a knowledge-base preset, we store its name, preferred terms and translations, replacement pairs, background notes, enabled state, timestamps, and account identifier so you can reuse it across devices. We also store service settings and state needed to provide features you choose, such as your saved-meeting preference and onboarding state. Privacy-consent choices for recording and AI summaries are stored locally on your device.

Billing and subscription data

If you purchase a subscription or minutes, RevenueCat manages product and entitlement information, Stripe processes web payments, and Apple processes Apple App Store purchases. We receive subscription status, remaining minutes, and a customer identifier. Payment card details are handled by Stripe or Apple and are not sent to Verli.

Referral rewards

When you use referral rewards, we store your referral code, the referral relationship, completed reward status and month, and notification state. The referral dashboard shows completed reward totals and notifications. It does not show referred identities or pending referral activity.

Usage and diagnostic data

We collect limited usage and diagnostic data through PostHog, Sentry, Axiom, and our hosting infrastructure to keep the Service reliable and understand which parts of the product are used. This may include IP address or a one-way hash of it for abuse prevention, browser or device type, operating system, account or device identifiers, pages you visit in the Verli web product, selected product-interaction events, crash reports, performance information, and structured request and application logs. Chrome extension analytics record app opens, successful interactive authentication, transcription starts, first translated results, stopped sessions, errors, exact session duration, language pairs, bounded product interactions, and extension version. They do not include selected-tab URLs, tab titles, raw audio, transcript text, credentials, arbitrary notes, or messages. PostHog may use cookies and browser or app storage; we disable autocapture and session recording.

Communications and privacy requests

For transactional messages, privacy requests, and content reports, we process the contact information and message content you submit, along with the report category, content type, non-content reference identifier, and timestamp needed to review and act on the request. Do not include private transcript or summary text in a content report. For product-education and lifecycle emails, we process your email address, first name when available, account identifier, preferred locale, platform and operating system, activation and product-usage milestones, subscription plan, signup status, and inactivity status.

How we use information

To provide real-time transcription and translation.
To authenticate you and keep your account secure.
To operate subscriptions, minutes, and promo codes.
To diagnose bugs and improve reliability.
To communicate important service notices.
To send product education and lifecycle emails after you activate the Service. You can unsubscribe from marketing emails at any time without affecting transactional or account-related messages.

Legal bases for processing

If you are in the European Economic Area, Switzerland, or the United Kingdom, we rely on the following legal bases:

Providing real-time transcription and translation, share sessions, saved meetings, and knowledge-base presets — performance of our contract with you. Where we show a recording disclosure before capture begins, we also rely on your consent to stream audio to our transcription processor.
Generating AI summaries and Read aloud speech — your consent.
Authenticating you, operating subscriptions and purchases, enforcing usage limits, and preventing abuse — performance of our contract and our legitimate interest in securing the Service.
Diagnostics, crash reports, product analytics, and acquisition attribution — our legitimate interest in operating and improving the Service. You can object to this at any time. On the website, storing analytics and attribution identifiers on your device also needs your consent where the law requires it (the European Union, the European Economic Area, the United Kingdom and Switzerland). You can withdraw it under Cookie settings at the bottom of every page.
Product-education and lifecycle emails — your consent, or our legitimate interest in communicating with existing customers; every such email includes an unsubscribe link. Transactional and account emails — performance of our contract.
Retaining tax and accounting records and responding to lawful requests — compliance with a legal obligation.

You can withdraw consent for recording and AI summaries at any time in Settings → Privacy permissions. These choices are stored on the device you are using, so withdrawing consent on one device does not change it on another.

Third parties we share with

We share the minimum data needed with the processors below. Where a processor handles personal data on our instruction, we put a data processing agreement in place requiring it to use the data only to provide the contracted service and to delete or return it when it is no longer needed. Some services listed here, such as content delivery networks, receive technical data such as your IP address simply because your browser connects to them to load part of a page.

Clerk — authentication and account management. If you choose Sign in with Google or Sign in with Apple, that provider processes the sign-in and shares your name and email address with Clerk.
Soniox — real-time speech-to-text and translation using captured audio, including audio from a browser tab you explicitly select in the Chrome extension, language settings, and optional context notes or preferred terms; and temporary batch processing of the compressed audio excerpt created by the public audio and video file translator. File-translator resources are deleted before the result is delivered.
MongoDB Atlas — database hosting for account-linked service data, including billing and usage records, share sessions and transcript segments, saved meetings and preferences, knowledge-base presets, desktop and Chrome extension authentication credentials, lifecycle-marketing records, acquisition-attribution records, content reports, and abuse-prevention counters keyed by your IP address or a one-way hash of it. A one-way hash of a deleted account identifier is kept so free-tier minutes cannot be reclaimed by re-registering. Historical Chrome guest-pilot metadata remains subject to its 90-day deletion period. MongoDB Atlas also temporarily stores public file-translator job metadata and bilingual text results for up to 24 hours. File-translator source media and raw audio are not stored in MongoDB Atlas.
Ably — real-time delivery of share-session transcript text and language settings to viewers.
RevenueCat, Stripe, and Apple — subscription and purchase processing. RevenueCat receives your account identifier, platform, operating system, and purchase events; Stripe receives your email address and payment details for purchases made on our website; Apple processes Apple App Store purchases. Windows purchases are made on the web.
Vercel — web hosting and request logs, including IP address. Verli API requests run on Vercel, including speaker voice clips while in transit.
Sentry, Axiom, and PostHog — error and crash reports, performance and structured request logs, and selected product analytics including email engagement, acquisition source, and purchase events. PostHog may use cookies and browser or app storage (on the website, only with your consent where the law requires it), is served from a Verli subdomain, and is also queried by us to attach acquisition information to internal signup notifications.
Google Gemini, xAI Grok, Lemonfox, and Google Cloud Text-to-Speech — optional transcript summaries, Read aloud speech generation, and translation of captions in the free YouTube transcript tool. If a selected speech provider is unavailable, another listed provider may be used.
Fly.io — hosting for Verli's speaker voice-classifier service, which processes short audio clips only for the duration of a classification request.
Plunk — product-education and lifecycle email delivery using your email address, first name when available, account identifier, preferred locale, platform and operating system, subscription plan, and events for activation, web activation, reaching the free limit, signup nudges, subscription starts, and inactivity. Plunk does not receive audio, transcripts, knowledge-base content, or meeting content.
Resend — transactional email delivery, including account welcome messages, privacy-request notifications and confirmations, and content-report notifications. Resend receives the recipient email address and the content and delivery metadata of those messages; a privacy request or content report may include the contact information, context, category, reference identifier, and timestamp you submit. Resend also delivers support and feedback messages you send us, and internal notifications to Verli staff about new signups (name, email address, and how you found Verli) and about purchases (email address and plan).
Cloudflare — bot protection on the free online tools and the contact form, which receives your IP address and browser signals; and hosting of desktop app downloads and update checks, which receive your IP address, app version, and platform.
Google Fonts — serves web fonts for our websites and emails, receiving your IP address and browser type when a page or email loads.
jsDelivr — delivers the in-browser audio-processing component used by the file translator, receiving your IP address and browser type.
YouTube (Google) and a third-party proxy provider — when you use the free YouTube transcript tool, we retrieve the captions of the video link you submit. The request is made by Verli through a proxy provider, not from your browser; the proxy receives the video link but no account information.

We require each processor that receives personal data on our instruction to provide the same or equivalent protection described in this policy.

We do not sell your personal information. We do not use your audio or transcripts to train machine-learning models.

How we secure information

Verli uses encrypted HTTPS and secure WebSocket connections when data moves between your device and hosted services. Account-linked data is stored with managed providers that protect stored data using encryption and access controls. Access to production systems and personal information is limited to authorized operators and service roles that need it to run or support the Service. Desktop refresh credentials are encrypted with Electron safeStorage, which uses operating-system protected storage such as Windows DPAPI or the macOS Keychain. We review access and delete or restrict data according to the retention and user-control commitments below. No system can be guaranteed completely secure, so contact support@verli.app if you believe your account or data has been compromised.

Retention

Account data is kept for as long as your account is active. When you delete your account we also delete your analytics profile and its events at PostHog. Error and crash reports at Sentry and structured logs at Axiom are not deleted individually; they age out on the retention schedule set with each provider. Marketing contact data is retained while your account is active or until you unsubscribe, and is deleted when your account is deleted. A minimal suppression record may be retained when necessary to honor an unsubscribe request. Share sessions and their transcript segments are deleted 24 hours after the session is created. Public audio and video file translator bilingual text results and job metadata are retained for no more than 24 hours; the source media is not retained by Verli, and Soniox resources are deleted before result delivery. The Chrome extension keeps unsaved transcript state only in its open app tab. Its rotating refresh credential is valid for up to 30 days and is removed from Chrome storage on sign-out, account deletion, or extension removal; its five-minute access credential stays in background-worker memory. Historical guest-pilot metadata remains subject to its 90-day retention limit, and the current signed-in extension does not create new guest sessions. Verli’s own billing and usage records are deleted when you delete your account; invoices and payment records held by Stripe, Apple, and RevenueCat are retained by them as required for tax and accounting purposes. Abuse-prevention records tied to your IP address, or a hash of it, expire automatically on a short schedule.

Saved meeting transcripts remain in your account until your account is deleted. Turning off Save meeting transcripts prevents future meetings from being saved but does not delete meetings already in History. You can request access, export, or deletion through the privacy-request and account-deletion links below.

Your privacy rights

You can access, correct, port, and delete the personal data we hold about you. Delete your account at any time via account deletion; submit other requests (or download a copy of your data) at privacy requests; or email support@verli.app. Depending on where you live, the rights below may also apply.

European Economic Area, Switzerland, and the United Kingdom (GDPR / UK GDPR)

You have the right to access, rectify, erase, restrict the processing of, and port your personal data, and to object to processing based on legitimate interest or direct marketing. Where processing is based on consent, you have the right to withdraw it. Our legal bases for each purpose are listed under “Legal bases for processing” above. You have the right to lodge a complaint with your local data-protection supervisory authority. We do not currently offer automated decision-making with legal or similarly significant effects. You can exercise your rights directly with us using the contact details above.

California (CCPA / CPRA)

California residents have the right to know what personal information we collect, the right to delete it, the right to correct it, the right to opt out of the sale or sharing of personal information, the right to limit use of sensitive personal information, and the right not to be discriminated against for exercising these rights. Verli does not sell personal information and does not share personal information for cross-context behavioral advertising. California residents under 18 may also request removal of content they posted; contact us as described above. The "Shine the Light" law (Civil Code §1798.83) gives California residents the right to request information about disclosure of personal information to third parties for direct-marketing purposes — we do not engage in such disclosures. The categories of personal information we collect, the sources we collect them from, the business purposes for each, and the categories of recipients are described in “Information we collect”, “Legal bases for processing”, and “Third parties” above; retention is described in “Retention”. We honor Global Privacy Control signals as an opt-out of sale or sharing; because we do not sell or share personal information, no further action is needed.

Brazil (LGPD)

Data subjects in Brazil have the right to confirmation of processing, access, correction of incomplete or inaccurate data, anonymization or deletion of unnecessary or excessive data, portability, information about entities with which we share data, and revocation of consent. Submit requests via the contact above.

Canada (PIPEDA)

Individuals in Canada have the right to access and challenge the accuracy of personal information we hold about them. Complaints may be directed to the Office of the Privacy Commissioner of Canada (priv.gc.ca).

Australia (Privacy Act / Australian Privacy Principles)

Australian residents have the right to access and seek correction of personal information. Complaints may be directed to the Office of the Australian Information Commissioner (oaic.gov.au).

To delete your account, see account deletion. To download a copy of your data or submit any other privacy request, see privacy requests. We aim to respond within 30 days for GDPR / UK GDPR / LGPD / PIPEDA / Australia requests and within 45 days for CCPA / CPRA requests, with shorter timelines where law requires. Because privacy requests on these pages are submitted while signed in, no further identity verification is needed.

If you are not a Verli user

If you are not a Verli user but your voice or words were processed by someone using Verli — for example as a meeting participant, or as a viewer of a shared transcript — the Verli user who started the session decides what is captured and is responsible for telling you and, where required, obtaining your consent. You can still contact us at support@verli.app to exercise your rights. We will ask for enough information to identify the data concerned, and we may need to involve the Verli user who created it.

Children

Verli is not directed to children under 13 (or the equivalent minimum age in your jurisdiction). We do not knowingly collect information from children.

International transfers

Your data may be processed in the United States and other countries where our processors operate. Where required, we rely on standard contractual clauses for cross-border transfers.

Changes to this policy

We may update this policy from time to time. Material changes will be posted on this page with an updated “Last updated” date.

Contact

Questions about this policy, or want to exercise your privacy rights? Email support@verli.app, or write to Inkvi LLC, 100 N Howard St Ste R, Spokane, WA 99201, United States.