What we collect
- · Account info: your email address (via magic-link sign-in).
- · Documents you upload, the text we extract from them, and the scores/briefs/chat messages generated from them.
- · Billing info handled entirely by Dodo Payments (we never see full card numbers).
- · Basic usage logs (timestamps, evaluation counts) for reliability and abuse prevention.
How we use it
To run the evaluation pipeline you request, show you your own past evaluations, enforce plan limits, and operate billing. We do not sell your data, and we do not show your uploaded documents or results to other users — except when you explicitly create a share link, which is intentionally public and read-only.
We never train models on your documents
Document text sent to the language model is used only to answer the specific request it was sent for (extraction, brief, or chat answer) and is not used to train or fine-tune any model, ours or our providers'. Public federal solicitations may be cached to return an identical score to a second user who uploads the exact same public document — this cache never mixes in or exposes any private customer data.
Where your data lives
Application data is stored in Supabase (Postgres), access-controlled by row-level security so each account's evaluations belong to that account. Documents are analyzed via LiteLLM, currently routed to Groq-hosted open models; the specific model provider is configurable and may change as the product matures. Payments are processed by Dodo Payments.
Retention and deletion
Evaluations are deleted automatically 30 days after creation by default. You can request deletion of your account and all associated data at any time by emailing support@zenexta.com.
Your rights
You can request a copy of your data, correction of inaccurate account information, or deletion of your account at any time.
Changes
We may update this policy as the product evolves. Material changes will be reflected by an updated date at the top of this page.