Edgee Launches On-Premise AI Gateway for Data-Sensitive Organizations

Summary

Edgee has announced the general availability of its AI Gateway as a fully on-premise deployment option, allowing organizations to manage AI model routing and observability entirely within their own infrastructure. The on-premise version provides the same advanced token compression, provider routing, and observability capabilities as Edgee's cloud offering, but runs completely within the customer's trusted environment—critical for organizations handling highly sensitive data such as financial transactions, clinical records, or government documents. Deployment is streamlined through Docker Compose for single-VM setups or Helm charts for Kubernetes 1.24+ clusters, with enterprise security hardening including non-root container execution and read-only root filesystems.

Edgee's on-premise solution supports two operational modes: connected mode, where configuration syncs every 15 seconds from Edgee's control plane while usage telemetry is reported back; and headless mode for fully air-gapped environments, where configuration is managed locally with zero outbound communication to Edgee. Critically, prompts, completions, and provider API keys never leave the customer's infrastructure in either mode. Organizations retain full control over bring-your-own-key (BYOK) credentials, spend limits, model allow-lists, and can route observability data to their own backends, maintaining complete compliance with data residency and security requirements.

• Enterprise deployment simplified through Docker Compose or Helm charts with security hardening (non-root containers, read-only filesystems) included
• Compliance and security teams gain transparency on data residency; platform teams get standard infrastructure tooling with versioned releases