Sub-processors

Last updated · August 2026

Every third party that can hold or see your data, what it does, and what it receives. Complete, and kept current with the code.

Who else can see your data

A sub-processor is any company that can hold or process your data because we asked it to. This is the complete list. If a company isn't here, it doesn't receive anything of yours.

Processors that hold your content

CompanyWhat it doesWhat it can see
AnthropicThe language model behind everything Lio reasons aboutYour chats, project context, retrieved abstracts, and the text of materials you add
OpenAIText embeddings for search; voice transcriptionText to be embedded; the audio you record when using voice input
MongoDBThe database everything is stored inAll of your stored content
VercelHosting for the app and the APIRequest metadata and server logs; content in transit
GumroadPaymentsYour purchase email. We never see card details — Gumroad handles those and we never receive them

Anthropic and OpenAI are called with our own API keys. Under their standard commercial API terms, what we send is not used to train their models and is held only for a limited abuse-monitoring window — 30 days at the time of writing. We have not signed a zero-retention agreement with either; that gap is stated plainly on our Privacy page.

Identity providers

Google and GitHub, and only when you choose to sign in with them. They tell us your name, email address and profile picture. They never tell us your password, and we never ask for one.

Google is not an AI provider for Theioptera. It appears here and nowhere else.

Public APIs we query on your behalf

These receive a short literature search phrase, sent anonymously from our server with no identifier of you or your account attached. They do not receive your project text, your chat history, or your materials.

  • OpenAlex
  • PubMed and the NCBI E-utilities
  • Semantic Scholar
  • Crossref / doi.org
  • bioRxiv and medRxiv
  • ClinicalTrials.gov
  • Reference sources: UniProt, Ensembl, Gene Ontology, UCSC Genome Browser

Separately, if you add a material by URL, our server fetches that page. The site sees a request from us, not from you — no cookies, no referrer, nothing identifying you.

What is not on this list

No analytics provider, no advertising network, no data broker, no session-replay tool, no CRM holding your research content. There is no analytics SDK in the product at all.

We will update this page before adding a sub-processor, not after. Questions, or a request to be notified of changes: yaraismail2611@gmail.com.

See also
Questions? yaraismail2611@gmail.com