Privacy Policy

Effective date: June 17, 2026  ·  App: Larken (iOS & Android)

Provider: Sumit Sehrawat (“Larken,” “we,” “us,” or “our”)  ·  larken.app@gmail.com

1Information We Handle

All of the information you put into Larken is information you choose to provide. Larken does not collect data about you in the background.

• Your background / résumé. Your name, work history, skills, and similar details — typed in, or imported from a PDF or a photo you select. This can include personal information (PII). It is stored locally on your device.
• Per-interview inputs. The job posting text, company information, and interviewer information (for example, a LinkedIn URL or your own notes) you provide for a given mock interview. You may type, import, or paste a public web link that Larken fetches for you (see §4).
• Your answers. Your spoken or typed responses. Spoken answers are converted to text on your device (see §3). The text is processed to generate coaching, then discarded — raw answer text is held only in memory during the round and is not written to long-term storage.
• Generated content. The questions, coaching, scores, and end-of-round “throughline.” A history of questions, coaching, and scores is saved locally (see §5).
• Subscription status. If you subscribe to Takeoff, Larken receives your entitlement status (active / not active) from Apple. Larken never receives or stores your name, card number, or any payment details — Apple processes all payments.
• Account & backup (optional). If you turn on Backup & sync, we store your email address (to sign you in via a one-time code) and an encrypted copy of your interview history on our servers, so it survives a lost or new phone and syncs across your devices. This is off by default and entirely optional — if you never enable it, we hold no account and no server-side copy of your content. See §5.

We do not collect: device advertising identifiers (IDFA), location, contacts, browsing history, or background usage analytics.

2How We Use Information

We use the information you provide only to operate the coaching experience:

• To generate mock-interview questions tailored to your background, the job, the company, and the interviewer.
• To evaluate your answers and produce coaching, scores, and your end-of-round report.
• To save your interview history so you can track progress over time.
• To provide and verify your subscription entitlement.

We do not use your information for advertising, we do not sell or share it for marketing, and we do not use it to build a profile of you across apps or websites.

3On-Device Processing (Voice, PDF, Links)

Several features run entirely on your device, so the underlying data never leaves it:

• Speech-to-text uses your device's on-device speech recognition. Your raw audio is never persisted and is never sent to a server for recognition. Only the resulting text is used, and only as described in §4. Where a device or language can't recognize on-device, voice is unavailable and you simply type instead.
• PDF text extraction (when you import a résumé/posting as a PDF) runs on-device.
• Reading a pasted link (e.g. a job posting URL) is fetched directly from your device — we don't route it through our servers.

The text produced by these on-device steps is then treated like any other text you typed in.

4Third Parties and Where Your Text Goes

To be honest and accurate: some of your text does leave your device.

a. Anthropic (AI provider). To generate questions and coaching, the relevant text content — your background, the job/company/interviewer inputs, and your answers — is sent over an encrypted HTTPS connection to the Anthropic Claude API. This is third-party processing performed by Anthropic, Inc. Per Anthropic's API terms, Anthropic does not train its models on API inputs or outputs by default. In our production architecture, these calls are routed through our own secure server proxy (so our API key is never embedded in the app); the proxy passes your text to Anthropic to generate a response and does not retain your content as a user profile. See Anthropic's privacy practices at anthropic.com/legal/privacy.

b. Apple & Google. Larken is distributed through the Apple App Store and Google Play, and subscriptions are sold and processed by Apple (StoreKit) or Google (Play Billing). Apple or Google handles your payment information under their own privacy policies. The operating-system frameworks Larken uses for on-device speech and document processing run on-device as described above.

c. Public web pages you ask us to fetch. If you paste a public URL for a job or company page, Larken makes a best-effort request to that page over HTTPS to extract its text for your mock interview. Larken blocks requests to private, internal, or loopback addresses. The website you point us to will see a normal web request from the app. Only paste links you're comfortable having the app retrieve.

d. Our backend (Cloudflare). Our secure server proxy — and, if you enable Backup & sync, the encrypted backup store — run on Cloudflare's infrastructure as our processor. Your backed-up history is encrypted at rest there; Cloudflare does not use it for its own purposes.

We do not use any analytics, attribution, advertising, or crash-reporting SDKs from third parties.

5Storage, Retention, and Deletion

• Local-only storage. Your résumé/background and first name are stored in the device Keychain (iOS) / encrypted storage (Android), set to this-device-only so they are not included in iCloud/Google backups. Your interview history (questions, coaching, and scores — not your raw answer text) is stored in an encrypted local file (AES-GCM-256, key held in the Keychain/Keystore). The app disables cloud backup of its data.
• Raw answers are not retained. The full text of your spoken or typed answers is held in memory only during a round and discarded afterward.
• Optional backup & sync. By default there is no account and no server-side copy of your content. If you opt in to Backup & sync (Profile → Backup & sync), your email address and an encrypted copy of your interview history are stored on our servers — encrypted at rest and transmitted only over HTTPS — so your history survives a device change and syncs across your devices. You can turn it off and delete the cloud copy at any time (Profile → Remove cloud backup), and signing out stops further backup. Your raw answer text and voice are still never uploaded.
• Retention. Content on your device stays until you delete it or uninstall the app. If you enable backup, the encrypted cloud copy is retained until you remove it or delete your account; the transient text sent to Anthropic exists only for the time needed to generate a response.
• How to delete your data. Open Profile → “Delete all my data” to permanently remove, in one step, your saved rounds, your background/résumé, your first name, your profile photo, and the encryption key from this device — and, if backup is on, the encrypted cloud copy and account too. You can also remove the app's local data by uninstalling Larken.

6Security

We design Larken to keep your information on your device and to minimize what leaves it.

• Content stored on-device is protected by OS sandboxing; interview history is additionally protected by device Data Protection (file encryption tied to your device passcode).
• Text sent to Anthropic is transmitted over encrypted HTTPS (TLS).
• Our API credentials are kept on our server, not embedded in the shipped app.

No method of storage or transmission is 100% secure, and we cannot guarantee absolute security. Keeping your device locked with a passcode and up to date materially improves the protection of your local data.

7Children's Privacy

Larken is not directed to children and is intended for users age 13 and older (and where a higher age of digital consent applies, that age). We do not knowingly collect personal information from children under 13. If you believe a child has provided us information, contact us at larken.app@gmail.com and we will address it.

8Your Rights and Choices

Because Larken stores your content locally and does not maintain accounts or server-side profiles, you are in direct control of your data — you can view, edit, and delete it on your device at any time, and uninstalling removes it.

For users in the EU/EEA and UK (GDPR/UK GDPR). Where applicable, you have rights to access, correct, delete, restrict, or object to processing of your personal data, and to data portability. The legal basis for processing your inputs is the performance of our service to you (and your consent for optional features such as microphone use). Much of this you can exercise yourself on-device; for anything involving our server proxy, contact us at larken.app@gmail.com. You may also lodge a complaint with your local supervisory authority.

For California residents (CCPA/CPRA). You have rights to know, delete, correct, and to opt out of “sale” or “sharing” of personal information. We do not sell or share your personal information (including no sharing for cross-context behavioral advertising), and we do not use it for targeted advertising. To exercise rights, contact larken.app@gmail.com. We will not discriminate against you for exercising your rights.

Permissions. Microphone and Speech Recognition access are optional — you can decline them and type your answers instead, and change these anytime in Settings. Photo and file imports are user-initiated and do not require broad photo-library access.

9International Users

Larken is operated from the United States. If you use the app from another country, the text sent to Anthropic may be processed on infrastructure located in the United States. By using Larken you understand that this processing occurs as described in this policy.

10Changes to This Policy

We may update this policy as Larken evolves (for example, if we add accounts or cloud sync). We will post the updated policy here with a new effective date and, for material changes, provide reasonable notice in the app. Your continued use after an update means you accept the revised policy.

11Contact

Sumit Sehrawat larken.app@gmail.com