One example, grown across three days

Rather than a disconnected exercise per topic, the programme grows a single realistic example that deepens each day — documents, then decisions, then actions, and finally a workflow that runs by itself. The scenario is set to your own domain during scoping, and Day 3 moves onto each participant’s real work.

What Each Participant Can Do Afterwards

Every outcome is something they will have built themselves, not watched someone demonstrate

  • Tell AI, generative AI, AI agents and agentic AI apart, using an example from their own work.
  • Use RAG to get grounded, cited answers from their own documents.
  • Write instructions that keep an AI agent inside its reference data instead of guessing.
  • Build and chain several AI agents that hand work to one another.
  • Make an automation write to a real system of record, not just print an answer.
  • Add branching so a workflow handles the exceptional case, not only the happy path.
  • Trigger a workflow on a schedule, a form, an email or a webhook — with nobody watching.
  • Put a human approval step in front of anything consequential.

Curriculum — Day by Day

Timed to an eight-hour day with 90 minutes of breaks — 390 minutes of instruction per day, roughly 75% hands-on. Concepts appear only where they are needed to make the next build work.

Day 1

Grounded answers — GenAI and RAG, hands-on

60 + 90 + 105 + 135 = 390 min
60MIN

Orientation, and the confident wrong answer

AI, generative AI, AI agents and agentic AI told apart on an example from your own operation, then why a fluent answer is not a correct one.

90MIN

Prompt and context engineering

Role, context, constraints, output shape and examples — turning a vague request into one that returns the same usable result every time.

105MIN

RAG — answers grounded in your own documents

Retrieval-Augmented Generation with no code, in NotebookLM: upload a document set, read the citations, see the honest “not in the sources” refusal, then add the missing document and watch the answer change.

135MIN

n8n foundations, and your first AI agent

Workflows, nodes, triggers and saved credentials. The AI agent node and the LLM behind it. Everyone builds a working single-purpose agent that refuses out-of-scope questions.

Day 2

AI agents that work together

90 + 120 + 105 + 75 = 390 min
90MIN

A second agent, and structured output

Why fixed fields matter the moment anything downstream consumes the answer, and how free text quietly corrupts a spreadsheet or a ticket queue.

120MIN

Agent orchestration — handing work along a chain

Chaining AI agents on one canvas and passing data between them, including the referencing details that catch everyone the first time.

105MIN

Conditional logic and branching

Conditions, filters and routing — urgent items down one path, routine down another. The biggest single difference between a demo and something a team relies on.

75MIN

Taking a real action

Connecting to a system of record and writing to it — access, field mapping, and why a durable log beats a notification email.

Day 3

Automation that runs by itself

90 + 90 + 105 + 105 = 390 min
90MIN

Triggers — starting without a human

Running on a schedule, on a submitted form, on an incoming email or message, or on a webhook from another system. The step that turns an assistant into an automation.

90MIN

Your own RAG pipeline inside the workflow

The grounded answering from Day 1, rebuilt inside n8n with a vector store, so it runs on your documents and your schedule inside your own automation.

105MIN

Approval, failure, cost and data safety

Human-in-the-loop approval before anything commits. Partial runs, duplicates and silent failures. Usage quotas and running cost. What must never be put into a hosted tool.

105MIN

Build your own, and know where to stop

Supported build time on a real task from each participant’s role, closing on the four lines — volume, auditability, latency and data sensitivity — where a workflow belongs with engineering.

Outcomes, Audience and Technologies

Each participant leaves with

  • A grounded RAG workspace built on their own team’s documents
  • A multi-agent workflow running end to end that writes a real record
  • One automation for their own recurring task, on a trigger, with approval and error handling
  • A written manual with every step and instruction, so it can all be rebuilt independently

The organisation gains

  • Automation capability outside the engineering queue — the people closest to a process can improve it
  • Working automations on day three, not a report recommending them
  • A shared understanding of what is safe to put into an AI tool and what is not
  • A clear boundary for escalating to engineering, and a handover they will accept

Who this is for

People who do not write code — operations, business analysis, finance, HR, customer support, quality, project management and team leadership. Comfort with a browser, email and spreadsheets is all that is assumed.

It suits organisations that want automation capability distributed across the business rather than concentrated in engineering.

Technologies and topics covered

Low-code AI · no-code automation · n8n · NotebookLM · RAG (Retrieval-Augmented Generation) · AI agents · agentic AI · LLMs (large language models) · GenAI (generative AI) · prompt engineering · context engineering · grounded answers and citations · hallucination control · structured output · agent orchestration · multi-agent workflows · vector store · document search · conditional logic and branching · webhooks · scheduled triggers · human-in-the-loop approval · error handling · API credentials · spreadsheet and system-of-record integration · workflow monitoring · usage quotas and cost · AI data safety and governance

Delivery

On-Site Corporate

  • Three consecutive days at your offices
  • Hands-on groups of five to six
  • A second facilitator for larger cohorts
  • Worldwide — Asia-Pacific, the Middle East, Europe and the Americas

Live Online

  • Six half-days across two to three weeks
  • Same content and same labs, with better attendance and retention
  • Breakout rooms for every lab
  • Any time zone

Practicalities

  • Nothing to install — everything runs in the browser
  • Delivered in English, paced for cohorts working in a second language
  • Labs run on your own documents and processes where access allows
  • A short readiness call and a simple setup checklist before delivery
Commercials on request

Frequently Asked Questions

Do participants need to know how to code?

No. The entire programme is low-code and no-code, built in n8n and NotebookLM in the browser. Comfort with a browser, email and spreadsheets is all that is assumed.

How many days is the low-code AI automation training?

Three days, timed to an eight-hour day with 90 minutes of breaks - 390 minutes of instruction per day, roughly 75% of it hands-on.

What is n8n, and why use it?

n8n is a visual workflow automation tool that runs in the browser. It lets non-programmers build AI agents, connect them to business systems, and run automations on a schedule or a trigger - without writing code or managing servers.

Does this cover RAG?

Yes. Day 1 covers retrieval-augmented generation hands-on in NotebookLM, so participants experience grounded, cited answers over their own documents. Day 3 rebuilds that as a RAG pipeline inside their own n8n workflow using a vector store.

Who should attend?

People who do not write code - operations, business analysis, finance, HR, customer support, quality, project management and team leadership. It suits organisations that want automation capability distributed across the business rather than concentrated in engineering.

Can it be delivered remotely?

Yes. Remote delivery runs as six half-days across two to three weeks rather than three consecutive full days - the same content and labs, with better attendance and retention.

What does each participant leave with?

A grounded RAG workspace on their own documents, a multi-agent workflow that writes a real record, one automation built for their own recurring task running on a trigger with approval and error handling, and a written manual to rebuild it all.

Related

RAG & Agentic AI Engineering (5 Days) →

The engineering counterpart: build AI agents and RAG pipelines in code, on LangGraph or AWS Bedrock.

All Training Programmes →

The full catalogue of AI and technical training, formats and how to book.

Corporate AI Training →

Team upskilling for engineering and L&D — build real in-house AI capability.

AI Workshops →

One-off hands-on workshops, from two hours to two days.

Bring This Training to Your Team

Three days, roughly 75% hands-on, no coding required. Download the full curriculum or book a free 30-minute scoping call.