Current posture stated plainly, including what LEVRYO does not have yet.
That depends on the design. A cloud deployment sends the specific data an agent needs to the model provider's API; a self-hosted deployment keeps orchestration inside your own infrastructure. LEVRYO sets this out explicitly during discovery, before any system is connected.
No. LEVRYO does not train or fine-tune models on client data. Where a model provider's API is used, the provider's own data-handling terms apply, and those are named during discovery so you can review them before anything connects.
Largely, yes. Self-hosted n8n keeps the orchestration layer on your servers. Calls to an external model provider still leave the network unless you run a model locally too, and that trade-off is made explicit rather than glossed over.
A data processing agreement can be put in place before any system is connected. Ask on the discovery call and it will be sent for your legal team to review.
LEVRYO does not currently hold SOC 2 or ISO 27001 certification, and will not imply otherwise. For buyers who require a certified processor today, we will say so rather than waste a procurement cycle.
The services that may process client data on LEVRYO's behalf are listed on this page. The list is maintained rather than assembled on request.
Access is revoked and working copies are removed as part of handover. Anything running on your own infrastructure stays with you, including the workflows and credentials, because they were yours throughout the engagement.
n8n is software, not a compliance certificate: compliance depends on deployment, where data rests, and what agreements are in place. A self-hosted deployment with the right controls can form part of a compliant workflow. LEVRYO will not claim more than that.