Back to home

Privacy Policy

Effective date: July 29, 2026

1. Introduction

Senvi ("we," "us," or "our") is a voice-first AI journaling application operated by Bisong Ventures, LLC. This Privacy Policy explains how we collect, use, disclose, and safeguard your information when you use our service at senvi.ai.

By using Senvi, you agree to the collection and use of information as described in this policy. If you do not agree, please do not use the service.

2. Information We Collect

Account Information

When you create an account, we collect your name, email address, and profile information through our authentication provider, Clerk. This includes any information you provide during sign-up or through a connected social login.

Journal Content

We store the text content of your journal entries, including messages you send during conversations, as well as AI-generated responses, themes, mood assessments, action items, and summaries derived from your entries.

Voice Data

When you use voice input, Senvi sends your audio to Deepgram, a third-party speech-to-text service, to convert it to text. On the web, your microphone audio is streamed to Deepgram; in our mobile app, a short recording is uploaded to Deepgram. If you upload an audio file, it is stored briefly while we transcribe it and deleted as soon as the transcription succeeds; if a transcription fails, the file is kept for up to 24 hours so you can retry without re-uploading, then deleted automatically. We send an opt-out flag on every request so that Deepgram does not retain your audio or use it to train its models. Apart from that transient upload handling, Senvi does not store your audio recordings; only the transcribed text is saved to your journal. Deepgram processes your audio on U.S. servers, encrypted in transit and at rest, under its own privacy policy and terms.

Spoken Replies

If you choose to listen to a reply, Senvi sends the text of that reply to Hume AI, a third-party text-to-speech service, to turn it into audio. This only happens when you tap to listen; we never send your reflections for speech synthesis on our own. The text we send is the coach's reply, not your own entries. The resulting audio is stored in your account so replaying a reply does not regenerate it, and it is deleted along with the rest of your data if you delete your account. Hume processes the text under its own privacy policy and terms.

Usage and Billing Data

We track entry counts and token usage for service operation and billing enforcement. If you subscribe to a paid plan, payment processing is handled by Polar (with Stripe as the underlying processor). We do not store your credit card number or full payment details on our servers.

3. How We Use Your Information

  • AI Processing:Your journal text is sent to the AI providers listed below to generate coaching responses and structured insights including themes, mood analysis, and action items. This processing happens server-side, in real time. None of these providers use your text to train their models: Anthropic's commercial terms forbid it, our OpenRouter requests are pinned to hosts that do not collect prompt data, Featherless does not log prompts or completions at all, and Fireworks does not train on API inputs or retain prompt data without an explicit opt-in we have not given.
  • Web Lookups: During a session the coach can look something up on the web. When it does, it sends a search query it composed from the conversation to Exa, our search provider. Exa states that it uses query data to train and fine-tune its own models, and offers no opt-out, so treat anything the coach searches for as leaving Senvi. Your entries themselves are never sent to Exa, only the query.
  • Service Operation: To provide, maintain, and improve the journaling experience, enforce usage limits, and manage your account.
  • Communication: To send service-related notices such as billing confirmations or important changes to the service.

4. Third-Party Services

We rely on the following third-party services to operate Senvi. Each has its own privacy policy governing how it handles data:

  • Clerk — Authentication and user management. Stores your account credentials and profile.
  • Convex — Backend database. Stores your journal entries, conversations, and account data.
  • Deepgram — Speech-to-text transcription. Receives your voice audio to produce a transcript. We opt out of its model-improvement program, so it does not retain your audio or use it to train its models.
  • Anthropic (Claude)— AI processing. Receives your journal text to generate insights. Under Anthropic's commercial API terms, Anthropic does not use API inputs or outputs to train its models and deletes them within 30 days.
  • OpenRouter — AI routing. Receives your journal text and forwards it to a model host. We pin every request to hosts that do not collect prompt data, so your text is not retained for training.
  • Featherless — AI model hosting. Receives your journal text. Featherless does not log prompts or completions; requests are processed in real time and not stored.
  • Exa — Web search. Receives search queries the coach composes during a session, never your entries. Exa uses query data to train its own models.
  • Fireworks AI — Text embeddings. Receives your message text to turn it into the numeric vectors that let the coach recall what you said before.
  • Resend — Email delivery. Receives the emails we send you, including recap emails, which contain content drawn from your reflections.
  • PostHog — Product analytics and session replay. Receives usage events and a recording of how you move through the app, tied to your account. Your reflections are masked out of these recordings.
  • LINE — Messaging, only if you connect a LINE account. Receives the messages exchanged in that channel.
  • Hume AI — Text-to-speech. Receives the text of a coach reply when you tap to listen, never your own entries.
  • Polar / Stripe — Payment processing for paid subscriptions. Handles billing details and payment transactions.
  • Vercel — Application hosting and content delivery.

5. Data Security

Your data is encrypted at rest and in transit (TLS). However, Senvi is not end-to-end encrypted. Your journal text is processed in plaintext on our servers and by the AI and infrastructure providers listed in section 4, which is what makes the coaching and recall possible. While we implement reasonable security measures, no method of electronic transmission or storage is 100% secure.

6. Data Retention

We retain your journal entries and account data for as long as your account is active, which is what lets your recap show progress over months. We do not delete your entries on our own, and we do not delete them if you downgrade to the free plan.

When you ask to delete your account, deletion is scheduled rather than immediate: there is a 30-day window in which you can cancel and get your account back. After that window your journal entries and personal data are purged from our systems. In the mobile app, signing back in during that window takes you to a screen offering to cancel; on the web you can still sign in and use the account normally until the window closes. Some data may persist in backups for a limited period before being permanently removed. You can start deletion from the account screen in the mobile app, or by emailing us at privacy@senvi.ai.

7. Your Rights

You have the right to:

  • Access your personal data stored in your account.
  • Delete your account and all associated journal data.
  • Export your journal entries (contact us to request an export).
  • Withdraw consent by discontinuing use of the service at any time.

To exercise any of these rights, contact us at privacy@senvi.ai.

8. Data Sharing

We do not sell, rent, or trade your personal data, and we will not start. We do not provide your data to third parties for advertising or marketing purposes, and we do not hand it to another company for that company's own purposes. We only share data with the third-party service providers listed above, solely as necessary to operate the service.

Your journal content is not used to train any AI model. The one exception is the web-search query described in section 3: when the coach looks something up, the query it composed reaches Exa, which trains on query data. Your entries are never sent.

We may use aggregated, de-identified information about how Senvi is used to improve Senvi itself: patterns across many accounts, not the content of your reflections.

Separately, and to be exact about it: the product analytics and session-replay provider listed in section 4 receives usage events and screen recordings that are linked to your account, so that data is neither anonymous nor confined to Senvi. Your reflections are masked out of those recordings, but the fact that you used the app, and how you moved through it, is visible to us there under your account.

9. Children's Privacy

Senvi is not intended for use by anyone under the age of 13. We do not knowingly collect personal information from children under 13. If we become aware that a child under 13 has provided us with personal data, we will take steps to delete that information.

10. Changes to This Policy

We may update this Privacy Policy from time to time. When we make changes, we will update the effective date at the top of this page. For significant changes, we will make reasonable efforts to notify you (for example, via email or a notice within the app). Your continued use of Senvi after any changes constitutes acceptance of the updated policy.

11. Contact Us

If you have any questions or concerns about this Privacy Policy or our data practices, contact us at:

privacy@senvi.ai