# Does Tessary use my data to train AI models?

No. Tessary runs no training or fine-tuning pipeline on customer data: your traces get read to evaluate your agent, never to train a model, yours or anyone else's.

Data also doesn't cross between organizations by default. Tessary runs single-tenant, so your traces, findings, and cases stay inside your own account with no pooling across customers. The one exception is opt-in: an org can choose to contribute aggregate statistics, never raw trace content, toward a shared baseline. Nothing crosses without that explicit choice, what crosses is a number rather than a trace, and it is not a training input even then.

---

Source: https://tessary.ai/answers/tessary-traces/does-tessary-train-ai-models-on-my-data
More on Tessary traces: https://tessary.ai/answers/tessary-traces
From Tessary, agent reliability for AI agents in production: https://tessary.ai