Deploy · Secure · Scale

Run one command.
Get a full private AI server.

A single Bash installer detects your hardware, configures Ollama, Open WebUI, and a secured reverse proxy, and hands you a production-ready AI infrastructure — no cloud, no subscriptions, no vendor lock-in.

Ubuntu · Debian · Rocky · AlmaLinux · Fedora · openSUSE x86_64 & ARM64
root@server: ~
The challenge

Private AI shouldn't require a systems team.

Most organizations want the productivity of AI tools without handing sensitive data to a third party — but standing up that infrastructure by hand takes days of manual configuration.

Common obstacles

  • Data privacy and compliance concerns with cloud AI providers
  • Recurring subscription costs that scale with usage
  • Vendor lock-in and unpredictable API changes
  • Complex, error-prone manual server configuration
  • Security hardening that gets skipped under deadline pressure

What the installer handles

  • Detects CPU, RAM, GPU, and storage automatically
  • Recommends the right AI models for the hardware present
  • Installs Ollama, Open WebUI, and a hardened reverse proxy
  • Runs every service under a dedicated non-root user
  • Sets up HTTPS, health checks, and automatic recovery
Under the hood

Everything a production AI server needs.

Automated installation

One script, start to finish — no manual package hunting.

Hardware detection

Reads CPU, RAM, GPU, and disk before it recommends a setup.

Enterprise-grade security

Non-root services, HTTPS, firewall rules, and Basic Auth by default.

Health checks & auto-recovery

A systemd timer restarts any service that goes down and logs the incident.

Automatic optimization

Model recommendations are tuned to the hardware actually present.

Reverse proxy & HTTPS

A single proxy terminates TLS and gates unauthenticated services.

Multi-distribution support

Ubuntu, Debian, Rocky, AlmaLinux, CentOS Stream, Fedora, openSUSE.

Update, repair, uninstall

The same command that installs it also maintains and removes it.

Complete AI stack

One installer, every module you need.

Chat

Multi-model conversations

Local chat with any supported Ollama model.

Coding

Coding assistant

Explain, refactor, and generate code offline.

RAG

Knowledge base

Document search grounded in your own files.

Analysis

Data analysis AI

Insights and reports from your datasets.

Image

Image generation

On-device image creation, no external API.

Video

Video generation

Text-to-video for supported hardware tiers.

Voice

Voice AI

Speech-to-text and text-to-speech, offline.

Gateway

AI API gateway

One endpoint routing to every local model.

Your infrastructure. Your data. Your rules.

Run the installer on any supported Linux distribution and you're done.

sudo bash install.sh
Get the installer