Privacy Policy

If you create an account we store your email address and, when you sign in through Google, the display name and avatar that provider returns. Nothing else about you is collected at sign-up.

When you run a model we store the job metadata — which model, when, how long it took, what it cost, and whether it succeeded. Job inputs are stored alongside that metadata so a run is reproducible.

Structure coordinates and other large outputs are kept in your browser's local storage, not on our servers. A predicted PDB file can be several megabytes and is always re-derivable from its source, so there is no reason for us to hold it.

Signed out, the platform works in full and everything stays local to your device.

We do not train models on your inputs, your outputs, or your research questions. We do not sell data. We do not share it with third parties for advertising.

When you run a model on a third-party provider — NVIDIA BioNeMo, Hugging Face, or your own Modal deployment — your inputs are transmitted to that provider to execute the job, and their terms apply to that transmission.

Queries to UniProt, the RCSB PDB, AlphaFold DB, ChEMBL, PubChem, ESM Atlas and Europe PMC are made server-side from Corollary. Those services receive the query text but not your identity.

Data is encrypted in transit and at rest. Credentials for external providers are read server-side only and never reach the browser.

Delete a job and it is removed from both your browser and the database. Delete your account and all associated rows are removed by cascade.

For any privacy question, or to request export or deletion of your data, contact us and we will respond within 30 days.