Voxily uses hosted infrastructure and AI generation, so here is the direct version: your data is processed only to draft the replies you ask for, and you review and send every single one.
Human-reviewed drafting inside the social composer. No auto-posting. These pages map each launch use case to the product behavior people can verify inside the Chrome extension.
When you request a draft — and only then — Voxily processes the visible post you are replying to, your typed draft, your selected mood, your persona settings, the knowledge entries you enabled, your account identifier, usage counters, and subscription status. Nothing is collected from pages you merely browse.
Voxily is hosted software — a browser extension, a backend API, and vetted providers — so we keep the promise simple: we process what is needed to draft the reply you asked for, and nothing more.
Voxily is a drafting assistant, not an automation bot. You review and send every reply yourself, and you must follow X/Twitter, LinkedIn, Reddit, and community rules. Voxily is built for authentic engagement — never for spam, deception, scraping, or automated posting.
Voxily is a drafting assistant. The user remains responsible for reviewing, editing, and posting the final text.
Voxily sends context to the backend when the user requests AI output, not just because a page is open.
AI drafts can be wrong or off-tone. The user edits and decides whether the content should be posted.
The extension does not auto-click post, send, like, follow, or message actions.
Users can clear local browser data and email support for account, persona, history, or subscription-related deletion requests.
Use these answers for launch comments, support replies, and early tester onboarding.
Yes. When a user generates a draft, Voxily routes the request through its backend and uses Groq for AI generation.
The product uses Convex for app data such as auth-related records, personas, knowledge text, history, usage, and subscription state where applicable. When you upload a PDF, DOC, DOCX, MD, or TXT for your knowledge base, the file is sent to our server once to extract its text, you review an AI summary before saving, and only the extracted text is persisted — the original file is discarded, never stored.
Voxily never reads your inbox in the background and never sends messages for you. When you ask for a DM draft, the visible conversation context near the composer may be included in that one request — the same way a reply works. You review and send every message yourself.
No tool can guarantee that. Voxily reduces risk by avoiding auto-posting and bulk automation, but users must follow platform and community rules.
Clear local extension data in Chrome and email support@getvoxily.com for deletion of account, persona, knowledge, history, or subscription-related records where legally possible.
Start with one real reply on X/Twitter, LinkedIn, or Reddit. Keep the draft only if it sounds like something you would actually post.
Install free