Built for the people who audit everyone else
Teddy works with your most sensitive compliance data. It runs in the region you choose, is read-only by default, never trains AI models on customer data and records every action an agent takes, so you can show your auditor exactly what happened.
How Teddy protects your data
Where your platform runs, and the principles that shape how Teddy and its agents work with your data.
You decide where your Teddy platform runs, so your data stays where your policies and customers need it.
Each customer’s data is logically separated. Agents only ever work inside your own workspace.
Data is encrypted in transit and at rest. Credentials for connected systems are stored encrypted and never shown in the interface.
Teddy does not train AI models on customer data. Your controls, policies and evidence stay yours.
Integrations use dedicated service accounts with read-only access. You approve the permission list before anything is connected.
Agents draft, people decide. Policies, risk acceptance and scope never change without a human approval.
Every agent action, sync and approval is recorded with time and actor, ready to show your auditor.
How agents get access, and how you stay in control
Agents only see what you connect, with the permissions you approve.
Security questions, answered
Where is our data hosted?
You choose where your platform runs: in the EU, in Germany or in the US. We set up your workspace in the region you select.
Does Teddy train AI models on our data?
No. Teddy does not train AI models on customer data. Your controls, policies and evidence stay yours.
Can agents change our systems?
Connections are read-only by default, and you approve every permission before a system is connected. Approved documents can be written back, for example to Confluence.
Who approves what agents produce?
People. Agents collect, check and draft. Policies, risk acceptance, scope and anything that goes out need a named approval.
Can our auditor see what the agents did?
Yes. The activity log records every agent action, sync and approval with time and actor, and you can export it for your auditor.
Talk to the team about your security requirements, or see Teddy in a demo.