Skip to content

Deployment Options

ByteFreezer offers two deployment models.

Deployment Compute Storage Control Plane Best For
On-Premises You You ByteFreezer Data sovereignty, compliance, most deployments
Enterprise Custom Custom Custom Air-gapped, custom integrations, dedicated support

On-Premises

You run all data-plane components (Proxy, Receiver, Piper, Packer, Query) in your infrastructure. Storage is your S3 or MinIO bucket. The control plane (configuration management, health monitoring, UI) is managed by ByteFreezer.

Your data never leaves your network. The control plane only stores metadata — dataset configs, health status, user accounts.

Install Methods

Use the installer project for ready-to-use deployment packages:

Platform Method Best For
Kubernetes Helm charts Production clusters (EKS, AKS, GKE, on-prem)
Docker Docker Compose Single-host, development, demos
Bare metal Ansible + systemd VMs, physical servers (Ubuntu/RHEL)
AWS ECS Fargate (CloudFormation + Terraform) Serverless on AWS
GCP GKE or Cloud Run (Terraform) Google Cloud
Azure AKS or Container Instances (Terraform) Microsoft Azure

Alternatively, the ByteFreezer MCP server can generate deployment packages via AI assistant — Docker Compose, Helm values, systemd scripts, or standalone scripts.

See On-Premises deployment for details.

Enterprise / Consulting

White-glove service for regulated or high-security environments:

  • Air-gapped deployments with no external connectivity
  • Custom integrations with your existing infrastructure
  • Dedicated support and onboarding
  • Compliance assistance (FedRAMP, SOC 2, HIPAA, PCI-DSS)

See Enterprise / Consulting for details.

Deploy with Claude + MCP

Skip manual setup entirely. Connect Claude Code to the ByteFreezer MCP server and describe your deployment in plain English. Claude creates tenants, datasets, generates configs, deploys via SSH or Helm, and verifies the pipeline.

See Deploy with Claude + MCP for instructions.

Demo Environment

A demo environment is available at bytefreezer.com for evaluation. Use it to test proxy configurations, explore the UI, and validate your data flow before deploying to your own infrastructure.