Agent Framework
n8n + AI
Business automation with AI built in
n8n is a visual no-code automation platform — think Zapier but self-hosted and more powerful. With AI nodes, your automations can make decisions, draft responses, categorize data, and more. Connect your email, calendar, CRM, databases, and APIs together with AI intelligence in the middle.
Book a consultation to set up n8n + AI →What it is
Automation that thinks
n8n is an automation platform you build visually — connect triggers to actions with a drag-and-drop editor. It supports 400+ integrations: email, Slack, Google Sheets, HubSpot, databases, webhooks, and virtually any API.
The difference from Zapier or Make: n8n runs on your hardware (no monthly subscription for flows, no data going to a US SaaS platform), and AI nodes let your automations make decisions instead of just moving data mechanically.
An email arrives. n8n classifies it using the local AI model. If it's a support request, it drafts a response and puts it in your drafts folder. If it's a lead, it creates a CRM entry and notifies you in Slack. All automatically, all locally.
How it works
Visual workflows, local AI
You build workflows in n8n's browser editor — drag nodes onto a canvas and connect them. AI nodes call your local Ollama model at any point in the flow: classify this text, summarize this email, generate this response.
n8n runs as a local service on your Mac. Workflows can be triggered by schedules, webhooks, or manual execution. The AI inference runs on Ollama — no data is sent to cloud AI services during automation execution.
Who it's for
Business owners who want automation without writing code
- ✓Business owners spending time on repetitive data tasks
- ✓Operations managers connecting systems that don't talk to each other
- ✓Small teams that want Zapier-level automation without the cloud costs
- ✓Anyone with email, CRM, or calendar workflows they want to automate
- ✓Businesses that handle client data and want automation kept local
Unique advantage
Self-hosted n8n has no per-task limits and no monthly automation fees. Your workflows run as often as needed without incrementing a counter. With local AI nodes, every decision stays on your hardware. No data going to Zapier's servers, no AI inference sent to OpenAI.
Full stack
What gets installed
| Layer | Component | Purpose |
|---|---|---|
| AI Engine | Ollama (MLX backend) | Runs models on Apple Silicon |
| Automation Platform | n8n | Visual workflow builder with 400+ integrations |
| AI Integration | n8n AI nodes | AI decision-making inside automation flows |
| Chat UI | Open WebUI | Browser chat, always available |
| Networking | Tailscale | Secure access from anywhere |
| Security | Hardened config | Self-hosted, no data sent to n8n cloud |
Security
Self-hosted, not cloud-hosted
We run n8n as a self-hosted instance on your Mac — not on n8n's cloud. Your workflow definitions, execution logs, and automation data stay on your hardware. AI inference runs through Ollama locally. We configure access controls and Tailscale for secure remote access to the n8n editor.
Recommended models
Models that pair well
Fast and lightweight — ideal for high-volume automation tasks like classification and routing
Very fast — works well for simple categorization, routing, and content generation in automation flows
Ready to set up n8n + AI?
Book a consultation. We'll map your workflows, configure n8n and the AI nodes on your Mac, and walk you through building your first automations.
Book a consultation to set up n8n + AI →