Trust & Security

When your client asks
"is this secure?" --
you'll have a clean answer.

Isolated per client

Every instance runs on its own server with its own filesystem, memory, and network. There is no shared database, no multi-tenant data layer, and no cross-client data access. One client's data never touches another's infrastructure.

You control the keys

API keys, email credentials, and third-party integrations are configured by you and stored on the instance itself. We don't access client data, and we don't hold copies of your keys. You own the relationship end to end.

Automatic updates

Security patches and runtime updates are applied automatically across your fleet. No version drift, no manual SSH sessions, no "I'll get to it next week." Every instance stays current without you lifting a finger.

AWS Infrastructure

Every instance runs on AWS, the same infrastructure trusted by banks, governments, and Fortune 500 companies. 99.99% uptime SLA, global data centers, enterprise-grade networking. Infrastructure you can depend on.

Infrastructure that earns trust

Each client instance runs inside a hardened systemd sandbox with read-only system partitions, private /tmp, and strict filesystem access controls. Network traffic is encrypted in transit via TLS, and data at rest lives on encrypted block storage.

Monitoring runs continuously across the fleet. If an instance goes silent, errors, or behaves unexpectedly, you're alerted before your client notices. Logs stay on-instance and are not aggregated to a central store unless you configure it.

No data leaves the instance except through integrations you've explicitly configured. There is no telemetry phone-home, no usage data collection, and no third-party analytics embedded in the runtime.

Security checklist

✓Per-instance server isolation
✓TLS encryption in transit
✓Encrypted block storage at rest
✓Hardened systemd sandboxing
✓No cross-instance data access
✓No telemetry or usage tracking
✓Automatic security patching
✓You own and control all API keys

What you can tell your clients

"Your assistant runs on a dedicated server that only you and I can access. Your data doesn't leave that server unless we've specifically configured an integration. I control the API keys, the permissions, and the update schedule. Nobody else has access -- not MoltBot, not the AI providers, not other clients."

That's the conversation you get to have. Not "we use enterprise-grade security" hand-waving, but a concrete, auditable answer about where data lives and who can see it. Your clients trust you. This infrastructure makes that trust easy to keep.

When your client asks"is this secure?" --you'll have a clean answer.