Deployment Options
Hosting options for the AutoEval platform
The AutoEval platform is available as a single-tenant hosted service from https://lastmileai.dev. It is free to get started, run evals, generate labels with LLM Judge, fine-tune custom metrics and create guardrails.
For customers needing more control over the platform data plane or control plane, often for added security and privacy guarantees, we offer VPC deployments on any cloud provider.
AutoEval is already in use at several Fortune 500 enterprises, and can be deployed within enterprise VPCs on AWS, Azure, GCP, and even on-prem. Please contact sales@lastmileai.dev to get in touch.
Service Architecture
LastMile AI provides deployable containers (OCI) that can be hosted on your private cloud through Docker, Podman, or any other OCI certified service.
Here is the AWS service architecture for the AutoEval platform. Similar deployments are available for Azure and Google Cloud.
Hybrid Deployments
In certain cases, you may want to have more fine-grained deployments (e.g. use the hosted platform for fine-tuning, but deploy the models and inference server in VPC). This hybrid deployment is also available, and depends on which part of the control or data plane you need control over.
Get Started with VPC Deployments
If you are interested in hosting AutoEval and alBERTa models within your private cloud, please contact sales@lastmileai.dev.