n8n
First met in aicp-c3-basecamp-03 · Automation platforms
n8nFirst met in aicp-c3-basecamp-03 · Automation platforms
Open-source workflow automation — nodes on a canvas, each node a small script, wired into if-this-then-that chains with optional AI agent nodes.
You want ownership and depth: self-host it free, customize node logic, and embed real AI agents — accept a steeper start than make.com/Zapier.
Fully open source and self-hostable — the only major automation platform you can run for $0 on your own box
AI agent node with model/memory/tools sockets — agents inside workflows, not bolted on
Workflows are portable JSON — generate, share, and debug them with a coding agent
Connects 400+ apps via prebuilt nodes (Gmail, Sheets, Firecrawl, WordPress...)
Runs AI agent nodes (model + memory + tools) inside workflows
Self-hosts free on your own machine or a cheap VPS; cloud version is paid
Imports/exports whole workflows as JSON — 'n8n is just a UI layer on top of JSON'
Doesn't think — deterministic pipelines need no AI node at all
Doesn't include the AI assistant on local installs (cloud only)
Doesn't replace a real backend for user-facing apps — it's plumbing, not product
The middle of your stack: triggers (schedule/webhook/chat) in, actions and AI processing in the middle, delivery (email/sheets/CMS) out. Reach for it when a task is a repeatable flow between existing services.
Automation platforms
Core course subject; not yet part of Paul's production stack — candidate for pipeline automations.
Could run the nightly course-archive checks if the Cowork scheduled task ever needs a self-hosted successor; the KB's scrape-and-publish patterns (BC4 newsletter) map directly onto future course-monitoring flows.
current
make.com: make.com is easier but closed-source, credit-priced, node-limited
zapier: same tier as make.com
langchain: n8n's AI features are built on LangChain under the hood (per BC2 trainer)
claude-code: coding agents generate/debug n8n workflow JSON — 'use AI to build AI'
| Session | Coverage | Moment | What happened |
|---|---|---|---|
| Basecamp 3: Introduction to n8n — I | demonstrated | 0:16:21 | Cloud version throughout: canvas, node panel, parameters, credentials, executions tab, AI assistant (build mode, @-tag nodes), JSON paste, template gallery, GitHub repo tour |
| Basecamp 4: Introduction to n8n — II | demonstrated | 0:26:50 | Full production build: schedule trigger, Firecrawl node install, scrape options, code nodes, AI agent, Google Docs, Gmail, sticky notes/colors, Ask-vs-Build AI debugging, JSON download/import, webhook variant |
| Basecamp 6: Voice Agents | demonstrated | 2:39:16 | Webhook → filter (call_analyzed) → Google Sheets append-row; drag-and-drop column mapping; expression for UNIX→readable date; outbound calls via Retell API key teased for the future session |
| Session 6: Building an AI-Powered Lead Gen Machine | demonstrated | 1:15:20 | The factory chassis: Google Sheets trigger, HTTP nodes (curl import), AI-written code nodes, Apify integration, LLM stations, sheet updates, sticky-note station labels; workflow JSON shared |
| Session 17: Facebook Ads Competitor Research & Replication | demonstrated | 0:52:49 | Entire nine-node pipeline built live on a 14-day free trial; UI recently updated with an AI assistant panel; Ask-AI used once for an HTTP body fix; auto-converted a hardcoded API token into a credential |
| Basecamp 5: MCP | explained | 0:19:15 | Callback anchor: the AI agent node (model/memory/tools) reframed as simplified tool calling; Gmail/Calendar nodes as prewritten API scripts |
| Session 2: How to Plan, Scope and Sell your AI Projects | explained | 1:34:22 | Build platform for client automations: largest integration catalog (its edge over Claude skills), the 500-page timeout war story, self-hosting for cost, weekly Google-Sheets re-auth reality |
| Session 12: Claude Code — Design & Building Agents (App-First: a Game, a GTM Board, and a Live Deploy) | explained | 1:54:57 | Q&A comparison: fixed automations vs Claude Code's AI automations; Claude can write n8n JSON workflows |
| Beginner Glossary (30 terms) + I'm Stuck Guides (23) | explained | Five glossary terms plus four dedicated diagnostic guides (vocabulary, credentials, Gmail/Sheets isolation, cost estimation) | |
| Practice Lab: Six Confidence-Building Projects | explained | Projects 3 & 4: label-scoped Gmail trigger, Edit Fields normalize, Code-node scoring, IF/Switch routing, webhook test events, Sheets lookup + append | |
| Learning Topics: Sessions 1–5 — Foundations to AI Director | explained | Session 4's entire automation vocabulary, canvas shortcuts, AI-node selection table, and the canonical support build | |
| Workbooks 1–5: Guided Builds with Completion Proof | explained | Workbook 5's platform: cloud trial path, credentials manager, the thirteen-step build, activation and monitoring | |
| AI Sprint: The Hermes Agent — Day 2 (Loop Engineering: the Scheduling Loop, the Delegation Loop, and the Skill Auditor) | explained | 'Is n8n gone? Strongly no' — the deterministic-pipeline lane, self-hosted for cost | |
| Session 1: Getting your first Client | mentioned | 0:20:40 | Cited as cohort skill inventory and lane-1 delivery vehicle (workflows, automations, dashboards) |
| Session 5: Mastering Lead Generation | mentioned | 2:30:56 | Tomorrow's automation: AnyMailFinder API + LinkedIn/website scraping + AI personalization + Instantly push, front to back |
| Session 11: Overview of Claude Code & Key Concepts (Installs, Modes, Context Files, Skills, Hooks & MCPs) | mentioned | 0:08:36 | His YouTube content subject; his agency builds these automations |
| Session 13: Claude Code — Skills, Sub Agents & the Token Economy (posted as 'Beyond Vibe Coding — n8n & More') | mentioned | 2:13:06 | An n8n skill exists in the marketplace — the title's only real n8n content |
| Session 15: Loop, Harness & Graph Engineering (posted as 'Securing Vibe-Coded Apps') | mentioned | 0:44:09 | The anti-example: fixed instructions, no self-adjustment — a workflow, not a loop |
| Session 16: GitHub 101 — Fundamentals & Essentials | mentioned | 0:42:39 | Scale exhibit: 1,607 branches, 22,000 commits on the public repo; its agent-skills repo shown as a fork-worthy asset |
| Session 18: Nano Banana + Veo 3.1 Multi-Shot Video — UGC Genie, an AI-UGC Product Built Live | mentioned | 'We're building our own n8n' — the dotted-node aesthetic and the is-n8n-still-relevant Q&A | |
| Session 19: Content Strategy Pipeline — the YouTube Growth Stack, Built by an Agent Team of Chats | mentioned | Skill-vs-MCP vocabulary from earlier weeks resurfacing in cohort answers | |
| Session 20: Complete LinkedIn Agency Automation — Carousels + Newsletter from One Content System | mentioned | 'Create an N8N-style visual workflow' — the named UI reference for the node graph | |
| Session 24: Building a Marketing OS — Sector-Agnostic Command Center on an Open-Source Stack | mentioned | The hosting analogy for PostHog-class self-hosted apps | |
| The Five-Level AI Generalist Roadmap (+ Operating System & Monetization) | mentioned | L4 automation platform shortlist with Make and Zapier | |
| AI Masterclass: Exploring Antigravity 2 — Four Products, Four Slash Commands, and Two Live Builds | mentioned | 'If you want the exact deterministic approach, use n8n' — the routing rule again | |
| AI Sprint: OpenClaw — Session 1 (Infrastructure, Not Agent: the Seven Documents, 1-Click VPS, and Model Routing) | mentioned | 'n8n is automations; OpenClaw is autonomous' — the comparison answer | |
| AI Sprint: OpenClaw — Session 2 (Token Anatomy, Claude Code Inside the Server, Skills, and Multi-Agent Workspaces) | mentioned | Slido positioning again: automation backend an agent can trigger, not an agent | |
| AI Sprint: The Open Source Playbook — Day 1 (Licenses, Open Weights, and Why Open Isn't Free) | mentioned | Case study: fully open source AND a paid-hosting business |