Everything to know before ordering.

What exactly do I get when I purchase a plan?

A working self-hosted AI stack on your own server: n8n for workflows, Ollama for models, and OpenWebUI for chat. Plus a PDF guide, admin credentials, a backup script, and a Loom walkthrough on most plans.

Do I need to provide my own server?

Yes. You provide the VPS, dedicated server, or local machine with SSH access. We can recommend specs during checkout if you are unsure what to provision.

What server specs do you recommend?

For 8B models: 16 GB RAM, 4 cores, 60 GB disk. For 13B–34B models: 32 GB RAM, 6+ cores, 100 GB disk. GPUs are optional but accelerate larger models significantly.

How long does setup take?

Quick Start delivers in about a day. Workflow Pack and RAG Pack take 3–4 business days. Complete builds take up to 5 business days, depending on the breadth of integrations.

Which open-source models can I pick?

Llama 3.1 (8B / 70B), Mistral 7B, Qwen 2.5 (7B / 14B / 32B), Gemma 2, Phi-3, and most other Ollama-compatible models. We help you pick based on your hardware.

Can you build custom n8n workflows for my use case?

Yes. Workflow Pack and Complete plans include hand-built workflows for your specific scenarios. Tell us in the order info form what you want automated.

Do you support RAG / chat with my documents?

Yes. RAG Pack and Complete plans include a vector store (Qdrant or Chroma), an ingestion pipeline, and chat-with-docs workflows wired to OpenWebUI.

Can I add or change models later?

Absolutely. Ollama makes it easy to pull new models. Your maintenance guide covers the steps for adding, updating, or removing models on your own.

Who owns the stack and the data?

You do. Everything runs on your server, under your control. We never retain access after handover unless you explicitly ask for ongoing support.

Is the stack secured by default?

Yes. Every install includes SSL via Caddy or Nginx, authenticated reverse proxy, firewall rules, and a credential rotation runbook.

How do you handle the SSH credentials I provide?

Credentials are encrypted at rest in your order record and accessible only to the engineer assigned to your build. We rotate or destroy them after handover at your request.

Is this a subscription?

No. Every plan is a one-time purchase. There is no recurring billing. Optional support extensions can be purchased separately.

What payment methods do you accept?

Visa and Mastercard, processed securely through Authorize.Net. We do not store full card numbers — they are tokenised at the moment of checkout.

Do you offer refunds?

Yes. We offer a 30-day money-back guarantee on every plan. See the Refund Policy for the full details.

How do we communicate during the build?

Every order has a built-in live chat between you and your assigned engineer. You will get notified the moment we send a message, and we get notified the moment you do.

What happens after delivery?

Your support window stays active for 7–60 days depending on plan. During that window we fix anything that breaks at no extra charge. After it closes, we still answer one-off questions reasonably.