Pipeline Monk
Practical guides on AI automation, local LLMs, and N8N workflows — written by someone who builds and runs this stuff every day.
What is Pipeline Monk?
Pipeline Monk is a technical blog for developers and solopreneurs who want to build real AI-powered systems — not just experiment with demos. The guides here cover running local LLMs on consumer hardware, building N8N automation pipelines, deploying AI agents, and everything in between.
Every article is written from hands-on experience. If something didn't work in testing, it doesn't make it into a guide. If a tool only runs well on a $10,000 server, that gets said clearly.
Who's behind it
I'm Kai Devlin, a full-stack developer who builds AI automation systems in production. I've been running local LLMs on my own hardware, building N8N workflows, and automating content pipelines long enough to know what actually works and what's marketing copy.
My daily driver for everything on this blog is a Ryzen 7 5800HS laptop with 16GB RAM and a GTX 1650. Not a cloud instance, not a workstation — consumer hardware you can actually buy. If a guide says something runs well, it means it ran well on that machine. If it was slow or borderline, the guide says that too.
I run Ollama locally with models like Qwen 3 7B and Phi-4 Mini. I build N8N pipelines that draft and publish content. I deploy on Cloudflare Pages. This site itself runs on the stack I write about.
A note on AI-generated content
Some articles on Pipeline Monk are drafted by an automated pipeline — Ollama + N8N doing the research and initial write. Those are credited to "AI Pipeline" in the byline. Every one of those posts is reviewed, fact-checked, and edited by me before it goes live.
Articles I write myself are credited to Kai Devlin. I believe in being transparent about the process — if AI helped write it, that's disclosed.
What we cover
- Local LLMs — Running Ollama, LM Studio, and other self-hosted inference tools on consumer hardware
- N8N Workflows — Building automation pipelines that connect AI models, APIs, and databases
- AI Agents — Tool-calling loops, multi-agent architectures, and production agent patterns
- Automation — Connecting AI to real business processes that run without you watching
- Tutorials — Step-by-step, command-by-command guides from setup to working system
- Tools & Reviews — Honest takes on what's worth using, tested locally, not based on benchmarks
Why trust Pipeline Monk?
Short answer: everything here has been tested on hardware I own, in workflows I actually run.
I don't write about tools I haven't used. Every model comparison is run locally on my machine. Every N8N workflow in a guide has been executed. I don't publish sponsored content disguised as reviews, and when I haven't tested something thoroughly enough to recommend it, the article says so.
The goal isn't to publish as much as possible — it's to publish things that are actually useful to someone trying to ship an AI-powered system.
Contact
Questions, feedback, or want to collaborate? Email works best: kaidevlin@pipelinemonk.com
You can also find me on GitHub: github.com/kai-devlin