The data Automedha handles, and why.

Short version

Automedha operates and publishes this site. We collect the information you submit for private-beta access or email updates. We do not run analytics, advertising pixels, or behavioral trackers, and we do not sell personal information.

Private-beta access requests

An access request includes your name, email address, company, use case, current stage, and timeline. We also store a generated reference, an idempotency key, a request fingerprint, the source domain, and timestamps. The fingerprint prevents one form key from creating conflicting requests; it is not a profile of your activity.

Automedha stores these records in its database before it attempts email delivery. The email service sends a confirmation to you and a notification to us. Email failure does not discard a stored request.

Email subscriptions

A subscription record contains your email address, its pending, confirmed, or unsubscribed state, and transition timestamps. Confirmation links expire seven days after issuance. We store Secure Hash Algorithm 256-bit token hashes instead of raw confirmation and unsubscribe tokens.

Subscriptions use double opt-in. An unsubscribed address must request and complete a new confirmation before it can return to the list.

Abuse prevention and rate limits

A bot-verification challenge checks access, subscription, search, answer, and voice transcription requests for automated abuse. Automedha sends the verification token and, when available, the connecting Internet Protocol address to the verification provider for validation. Verification tokens are single-use and expire after five minutes.

Location-based rate limiting uses the connecting address to enforce request limits. Automedha does not add that address to access-request or subscriber records in its database.

Search, answers, and voice

Ask Automedha sends your question to a hosted AI model and searches the public Automedha corpus with a keyword index built into the site and a vector index. Answer generation receives retrieved public excerpts so it can cite its sources. An AI gateway applies the service’s authentication and spending controls.

Voice mode sends accepted audio to a hosted AI model for transcription. Automedha does not store the uploaded audio in its database or vector index. A short-lived, actor-bound grant authorizes speech generation for the resulting text.

Automedha does not use access requests, subscriptions, questions, answer context, or voice audio to train models.

Infrastructure services

The site runs on a managed serverless runtime and uses a database, a vector index, hosted AI models, an AI gateway, bot verification, email delivery, rate limiting, and related infrastructure. These services process data needed to deliver and protect the requested feature. They are provided by a single infrastructure vendor, which handles that processing under its service terms and privacy commitments.

Browser storage and tracking

The access form keeps a request key in your browser’s session storage so a retried submission does not create a duplicate request; it is cleared when the session ends and identifies the request, not you. Ask Automedha keeps nothing in your browser. The site has no analytics or advertising trackers.

Access, correction, and deletion

You can ask for a copy, correction, or deletion of data connected to your email address. We may retain information when law, fraud prevention, or an active dispute requires it. Send requests to privacy@automedha.ai.