Security & Privacy

Every Sarudo client gets their own server. Your AI employee runs on your own virtual private server with its own database, its own file storage, and API keys held in provider accounts registered to you. There is no multi-tenant architecture — your data never touches another client's data. This is fundamentally different from most SaaS products, where every customer shares one database. Your instance is completely isolated, and it is yours: you hold the provider account, you get the bill, and nothing about it depends on a contract with Sarudo continuing.

You own the instance, the infrastructure it runs on, the accounts it works through, and the data inside it. The server, the database, the file storage, and the provider accounts are all in your name and billed to you directly — the knowledge entries, CRM records, documents, and conversation history stored there are yours outright. Your data is never used to train AI models for other clients. Nothing is held hostage by a subscription, because there is no subscription: stop buying support hours and your AI employee carries on running on your own infrastructure. You can request a full export at any time. We do not sell, share, or monetize your information in any way.

Traffic between you and your server is encrypted in transit with TLS. Data on the server itself is not encrypted at rest — it sits on the provider's standard disk, and your credentials live in a configuration file on that server rather than an encrypted store. Your instance runs on a server no other client shares, so your data is not co-mingled with anyone else's. Server access is restricted to key-based SSH authentication with no password login. A full backup is taken daily, verified readable, checksummed, and written with owner-only permissions to that same server — which protects against accidental deletion and corruption, but not against losing the host. If you need encryption at rest or an off-site copy, ask and we will scope it.

Sarudo implements mandatory approval gates for any action that has real-world consequences. Before your AI employee sends an email, makes a phone call, processes a payment, or publishes a social media post, it will present the action for your review and ask for explicit approval. You can approve, deny, or request modifications. This ensures you maintain full control over all external communications and financial transactions.

Your AI employee will never send an email without your approval. It will never make a phone call without your confirmation. It will never process a payment without explicit authorization. It will never post to social media without your review. It will never share your data with third parties. It will never use your data to train models for other clients. These are hard rules built into the system, not just policies — they are enforced at the code level.