Question 1

What is Enclo?

Accepted Answer

Enclo is a specialist engineering firm that builds and then operates private AI on infrastructure its customers control — their datacenters, colos, or their own cloud accounts. Offerings: private LLM stacks (open-weight models on vLLM), agent-ready platforms, and custom model serving. A fixed-fee deployment gets the platform standing up; the monthly operations retainer — 24/7 monitoring, upgrades, and incident response — is the ongoing engagement.

Question 2

Does Enclo host customer workloads?

Accepted Answer

No. Everything runs on hardware the customer owns or in cloud accounts the customer controls. Enclo never runs customer workloads on GPUs it owns or rents, and customer data — prompts, outputs, documents, model weights — never transits Enclo's systems.

Question 3

Who does Enclo serve?

Accepted Answer

AI software vendors whose enterprise customers require on-prem or VPC deployment of the vendor's product; AI companies and compute providers that need an inference platform or control plane operated reliably; and enterprises that want AI on their own GPUs or cloud tenancy — because their data can't leave, because API costs are scaling with headcount, or because hardware they bought is under-used.

Question 4

Does Enclo work in the cloud or only on-premises?

Accepted Answer

Both, with one requirement: the infrastructure must be under the customer's control. That means the customer's datacenter or colo (including fully air-gapped sites), the customer's own AWS, Azure, GCP, or OCI accounts, or a hybrid of the two.

Question 5

Does Enclo build AI agents or applications?

Accepted Answer

No — that's a deliberate line. Enclo builds and operates the platform agents run on (model serving, tool and MCP gateways, sandboxing, permissions, observability, evals). The customer's developers or application partners build the agents and applications on top.