Frequently Asked Questions

All data is fully encrypted and stored on European servers. Encryption keys are hosted with a separate European cloud provider and are never shared with our hosting platform, ensuring complete separation of concerns. Data can only be decrypted by authorised users through the platform, with separate keys per organisation.

All data is hosted on EU-based servers, fully compliant with GDPR requirements. Encryption keys are managed by a separate European cloud provider, ensuring that no single vendor has access to both your data and the keys to decrypt it.

No. AI conversation content and filesystem information are never accessible to the Svalin agent. Only the input and output of MCP server tool calls are stored and logged. Your teams' prompts, responses, and local files remain completely private.

Unlike traditional monitoring or MDM tools, Svalin is purpose-built for AI governance. It monitors and enforces policies across every MCP server, regardless of which AI client is used or whether the MCP server runs locally or remotely via API. This gives you a single governance layer across your entire AI stack.

Svalin works with any client that supports MCP servers, including terminal-based tools and desktop applications. As the MCP ecosystem grows, Svalin automatically supports new clients without any additional configuration.

Zero impact on your teams. Local agents simply log information and never interfere with actual tool calls. Your teams continue working uninterrupted — no blocking dependency, no downtime risk.

Days, not months. Deployment time depends on your client setup (via MDM or other distribution methods), but you can start defining policies and rolling out local agents to your teams right away.

Yes. Svalin supports granular policy configuration, allowing you to define different rules for different teams, departments, or use cases within your organisation.