AI
AI automation with n8n
Automate your business workflows with n8n. Process orchestration, multi-system integration, AI agents, RAG, the MCP protocol, real cases.
By the end of the course, participants will be able to:
- build automation workflows with n8n
- bring LLMs (large language models) and AI agents into your automations
- implement RAG (retrieval-augmented generation) solutions
- develop custom AI assistants
- automate the processing of documents, emails and structured data
- master the concepts of prompting, embeddings and vector databases
- deploy interactive AI chatbots and automated monitoring systems
Who it's for: developers, business analysts, IT professionals, automation specialists.
Prerequisites: a grounding in REST APIs, familiarity with JavaScript (basics).
Day 1 · Module 1 — Introduction to n8n and core concepts
- Introduction to n8n: features, use cases
- Workflow architecture: nodes, data links, data types (JSON, string, etc.)
- Flow logic: merge, split, loops, conditions
- Authentication methods (API keys, OAuth, etc.)
- Comparing editions: Community, Pro, Enterprise
Labs
- Building a simple workflow: retrieving data through a REST API
- Handling JSON data and linking nodes together
- Exercise on flow logic: loops and conditions, Merge
Day 1 · Module 2 — JavaScript in n8n and user management
- JavaScript basics for n8n
- User and permission management
- Self-hosting n8n: configuration and good practice
Labs
- Writing custom JavaScript functions in n8n
- Exercise: transforming data with JavaScript
Day 2 · Module 3 — NLP, LLMs and OpenAI integration
- NLP (natural language processing) and LLM concepts
- LLM parameters: temperature, tokens, etc.
- Introduction to LangChain and agent chains
Labs
- Building an AI chatbot to work with RSS feeds
- Exercise: automatic summarisation of Gmail emails with OpenAI
Day 2 · Module 4 — Email automation and AI assistants
- OpenAI Assistants API
- Gmail integration and knowledge base management
- Customising automated responses
Labs
- Configuring an AI assistant to process Gmail emails
- Generating contextual responses and integrating product knowledge bases
Day 3 · Module 5 — RAG and vector databases
- RAG (retrieval-augmented generation) architecture
- Vector databases (Pinecone, Weaviate)
- Processing PDFs and documentation
Labs
- Setting up a RAG system to index PDF documents
- Configuring Pinecone and integrating it with OpenAI
Day 3 · Module 6 — Document chatbots and deployment
- Building specialised chatbots
- Prompt engineering: good practice
- Embedding chatbots into a website through n8n
Labs
- Developing an interactive chatbot for Google Sheets
- Exercise: embedding chat into a website
Day 3 · Module 7 — Advanced integration with MCP (Modular Context Protocol)
- Introduction to the MCP protocol: vision, goals, principles
- MCP architecture: the roles of the client, the server and dynamic contexts
- Differences between MCP and traditional agents (agent chains, LangChain tools)
- What MCP brings to AI systems: scalability, modularity, interoperability
- Real MCP implementations: Brave Search MCP (contextual enrichment of queries), FireCrawl MCP (goal-oriented crawling with context injection), Claude Desktop MCP (user context integration through a custom MCP server)
Labs
- Deploying a custom MCP server in n8n (MCP Server Trigger, client-side configuration)
- Integrating an n8n flow with Brave Search MCP to enrich web queries
- Building a micro-agent with FireCrawl MCP for contextual scraping
- Connecting an AI agent system to a personal MCP server to receive context-enriched instructions
Day 3 · Module 8 — Production and optimisation
- Production architecture with n8n
- Integrating external modules
- Optimising workflows
Labs
- Deploying a complete workflow: from design to production
- Exercise: automating a business process with n8n and AI
Quoted on request — reply within 48h