> ## Documentation Index
> Fetch the complete documentation index at: https://docs.galagos.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Data privacy and security on Galagos

> How Galagos protects your research data: zero-retention AI, sandbox isolation, EU-only infrastructure, and TLS encryption.

Galagos is built for research teams who handle sensitive biological and health data. Every layer of the platform — from how AI requests are routed to where your files are stored — is designed so your data stays private, stays in the EU, and is never used to train AI models.

> [!WARNING]
>
> Galagos is not a medical device and does not provide medical, diagnostic, or regulatory advice. All results are for research purposes only and must be independently validated before any clinical or regulatory use.

## Zero-retention AI

When you run an analysis, Galagos routes your request to enterprise-tier LLM APIs that operate under strict zero-retention agreements. Your data is never used to train or fine-tune any AI model — by Galagos or by its LLM providers.

All requests are encrypted before transmission and routed only to the providers listed in the [Data Processing Agreement](https://docs.galagos.ai/security/gdpr-and-dpa.md).

## Sandbox isolation

Every Galagos project runs in its own isolated, containerized environment. Your files, intermediate outputs, and analysis state cannot be read by, or leak into, any other project — including other projects within your own organisation. When a project environment is torn down, its filesystem is discarded.

## EU-only infrastructure

All data you upload to Galagos is stored and processed exclusively within the European Union. Galagos runs on dedicated, single-tenant root servers operated by Hetzner in Germany and Finland. Data never transits outside the EU.

> [!NOTE]
>
> Dedicated infrastructure means your data does not share physical or virtualised resources with other Galagos customers.

## Encryption in transit

All data transmitted between your browser and Galagos, and between Galagos and its LLM providers, is encrypted using TLS. There is no unencrypted path for data in motion.

## LLM providers

Galagos uses the following LLM providers. Each has been selected for its data handling guarantees and GDPR compliance posture.

| Provider | Processing location | Data handling |
| - | - | - |
| Google Cloud Vertex AI | EU | EU-only inference; data processed within the EU |
| OpenAI | EU inference only | Zero-retention API; Standard Contractual Clauses (SCCs) |
| Anthropic | EU inference only | Zero-retention API; Standard Contractual Clauses (SCCs) |

You can also connect your own LLM provider. See [Bring Your Own LLM](https://docs.galagos.ai/security/bring-your-own-llm.md) for details.

## Your responsibilities

Galagos processes data on your behalf as a Data Processor. Before uploading any data, you are responsible for the following:

> [!WARNING]
>
> Do not upload biological, genetic, or health data that contains direct personal identifiers — such as names, standard dates of birth, or contact information. All such data must be strictly pseudonymized or anonymized before upload.

* **Pseudonymize or anonymize all biological and health data** before uploading. Remove names, dates of birth, contact details, and any other direct identifiers.
* **Establish a valid legal basis** under GDPR Article 6 and, where applicable, Article 9 before submitting any personal data.
* **Provide required privacy notices** to data subjects and obtain any necessary consents.

## Further reading

> [card] [GDPR and Data Processing Agreement](https://docs.galagos.ai/security/gdpr-and-dpa.md)
>
> Review your obligations as Data Controller and Galagos's formal DPA commitments.

> [card] [Bring Your Own LLM](https://docs.galagos.ai/security/bring-your-own-llm.md)
>
> Connect your own LLM provider or deploy Galagos entirely on-premise.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.