Privacy Policy
Interpreter Pro · Effective 15 April 2026 · Last updated 15 April 2026
Interpreter Pro ("the extension", "we", "our") is a Chrome browser extension that helps professional interpreters transcribe audio and translate it in real time. This document describes what data the extension processes, who it shares data with, and what control you have over it.
This is a plain-English document, not a legal contract. We've tried to write it the way a careful engineer would explain it to a colleague.
TL;DR
- We capture audio from your microphone or browser tab only when you press Record, send it to OpenAI Whisper for transcription and OpenAI GPT-4o mini for translation, and show you the result.
- We do not have user accounts. We do not store your name, email, phone number, or any personally-identifying information on our servers.
- Transcripts are stored on your device, in your Chrome browser's local storage. We never receive them.
- We do store a small monthly request counter on our server (Cloudflare Workers KV) tied to a random UUID generated when you install the extension. That's how we enforce the free-tier quota.
- Standard tier compliance scope: Interpreter Pro's standard tier (Free + Pro) is designed for general professional interpreting and is not currently configured for HIPAA-regulated workflows. For interpreters working with Protected Health Information (PHI), our Healthcare tier with full Business Associate Agreement (BAA) support is in development — join the early-access waitlist.
What we collect, and where it goes
Audio (transient)
When you press Record, the extension captures audio from the source you selected (your microphone, your active browser tab, or both). When you press Stop, the captured audio is sent to:
- OpenAI Whisper (via OpenAI's
/v1/audio/transcriptionsAPI) — primary transcription provider. - Groq Whisper (via Groq's API) — fallback provider, used only if OpenAI is unavailable.
The audio transits our Cloudflare Worker (a stateless routing layer with no disk storage) on its way to OpenAI or Groq. The audio file itself is never written to our storage. The Worker holds it in memory only long enough to forward the request.
Audio capture stops the instant you press Stop or close the side panel. The extension does not record in the background.
Transcripts and translations (local-only)
The transcribed text and translated text are returned to your browser and stored in chrome.storage.local on your device. They appear in the side panel transcript area and persist across browser sessions until you click the Clear transcript button or uninstall the extension.
We never receive your transcripts on our servers. They live entirely on your computer.
Free-tier quota counter (server-side)
On first install, the extension generates a random UUID and registers it with our server (a POST /register call). We store a per-month counter in Cloudflare Workers KV with a key like quota:free_<uuid>:2026-04 and a value like 23. This counter is used to enforce the 50-requests-per-month free-tier limit. The counter is automatically deleted 35 days after the month ends.
We do not link this UUID to any identifying information. We do not log the content of your requests — only the count.
Settings (local-only)
Your language pair, audio source, and microphone selection are stored in chrome.storage.local on your device. We don't see them.
Network metadata (Cloudflare-side)
When you make a request, Cloudflare records standard request metadata (IP address, timestamp, response status, user agent) for routing and abuse prevention. This is governed by Cloudflare's privacy policy, not ours.
Sub-processors
| Sub-processor | What they receive | Purpose | BAA / DPA |
|---|---|---|---|
| OpenAI (privacy) | Audio + transcribed text via API | Primary transcription and translation | No BAA. By default, OpenAI does not train on API data (as of March 2023). |
| Groq (privacy) | Audio + text via API | Fallback transcription and translation when OpenAI is unavailable | No BAA. |
| Cloudflare (privacy) | Request metadata, free-tier UUID, monthly counter | Hosts our Worker and KV storage | Standard Cloudflare terms only. |
| LemonSqueezy (privacy) | Email, name, payment info — only if you upgrade to paid tier | Handles billing for the Pro subscription | Standard LemonSqueezy terms. The extension only receives a license key, never your billing details. |
What we do NOT collect
- Email addresses (unless you upgrade to paid tier, in which case LemonSqueezy collects it for billing — we never see it)
- Names, phone numbers, addresses
- Browsing history
- Tab content beyond what you explicitly choose to share when you select "Call Audio"
- Microphone audio when you are not actively recording
- Data from third-party tabs or other extensions
- Diagnostic telemetry, analytics, or crash reports
Your rights
- Delete your local data: Click Clear transcript in the settings panel, or uninstall the extension entirely. Either action wipes everything we store on your device.
- Reset your free-tier counter: Clear the extension's
chrome.storage.sync(right-click the extension → Manage extension → Site data, or via Chrome's storage settings). A new UUID will be generated on the next request, giving you a fresh monthly quota. We technically cannot prevent this, but please use it sparingly — it's how we keep the free tier sustainable. - Request data deletion from sub-processors: Contact OpenAI, Groq, or Cloudflare directly using their published procedures. We have no special access to their stored data on your behalf.
- Opt out of OpenAI / Groq training: Both providers offer API-data opt-outs. By default, neither uses API data for model training as of the dates listed in their respective policies.
Compliance disclosures
- GDPR: We process minimal data (a UUID and a counter). We are not established in the EU and have not signed bespoke Data Processing Agreements with our sub-processors beyond their default terms. EU users should consider this before use.
- Children: The extension is not intended for users under 13. We do not knowingly collect data from children.
- Healthcare regulation (HITRUST, etc.): Not certified.
- Financial regulation (PCI DSS): Not in scope. We do not process payment data ourselves; LemonSqueezy handles all payment processing for paid-tier subscriptions.
- Court / legal proceedings: Not validated for legal admissibility. Transcripts produced by Whisper are subject to transcription errors and should be verified by a human before any use in legal proceedings.
Changes to this policy
We may update this policy when our data practices change. The "Last updated" date at the top will reflect the latest revision. Material changes will be announced in the extension's release notes on the Chrome Web Store.
Contact
For privacy questions or data requests, please email: