Building a Command Centre with Claude/Part 1

It’s been a while since I made the bold decision to keep pace with AI.

I didn’t want to be just an observer. I decided to be an active learner.

My focus lately has been on building hands-on fluency with automated reporting and agentic workflows using LLM-powered tools like Claude.

As a birthday gift, I received six months of Claude Pro, and it has become my learning ground for exploring how Claude Cowork and Claude Code work together.

I have to admit that finding real use cases wasn’t straightforward at first, as testing these tools on a personal computer was making it hard to simulate an organisational environment.

But gradually, Claude helped me generate the files and my first scenario that made the use case tangible and transferable.

Next month, I’ll share the result. This month, I’m sharing the process and the thinking behind it.

There is a particular kind of courage in being a beginner again.

Especially when you have spent years being the person who knows. Who explains. Who guides others.

Learning AI tools has reminded me that mastery is not a destination.

One of the most quietly radical shifts that AI is creating for professionals is this:

The skill that matters most is no longer knowing the answer. It is knowing how to design the question.

When I started exploring Claude Cowork and Claude Code together, I wasn’t just testing features, I was learning to think in systems.

This is a skill that sits at the intersection of operational excellence and leadership.

A Chief of Staff who can tell you the health of the business at 8am on Monday, without spending Sunday night compiling spreadsheets, is not just more efficient.

The Command Centre I am building is a small-scale version of that vision. It is an agentic system that collects, processes, interprets, and delivers on a schedule, without manual input.

And the surprising part?

The hardest work was not the code. As Claude knows coding. It was the clarity. Deciding what mattered. Defining the signals worth tracking. Writing prompts precise enough to produce an executive summary worth reading.

That is the new operational skill. Not just working with AI, but directing it well.

Imagine you are the Chief of Staff of a growing organisation. Every Monday morning, the CEO expects a clear picture of where the business stands, across Growth, Delivery, Finance, and Talent.

Today, that report is assembled manually: emails chased, spreadsheets refreshed, slides updated.

Here is what the same workflow looks like with an agentic system built in Claude:

What it tracks:

• Growth — New opportunities this week · Pipeline value (£) · Win rate (%)

• Delivery — Active projects · Team utilisation (%) · Projects at risk

• Talent — Open roles · Time to hire (days) · Offers accepted this week

• Finance — Revenue vs target (%) · Cash runway (weeks) · Overdue invoices (£)

What it does:

The system collects data from simulated function inputs across Growth, Finance, Delivery, and Hiring, passes them through the Claude API to generate a structured executive summary, and delivers it automatically every week, via email, alongside a simple, always-updated dashboard.

Every Monday at 8am, the CEO receives a digest they can act on. No one had to build it by hand.

This is the use case I am building. The result, including what it actually produces, will be in upcoming editions.

If you want to build your own version, here is the learning path I followed.

Block 1: Build the Foundation

• Set up an n8n Cloud account and complete your first workflow: a manual trigger that sends an email. Just that. The goal is to get comfortable with the environment.

• Create a Google Sheet with simulated data: Finance, Growth, Delivery, Hiring. Fill in four weeks of invented numbers. This becomes your data source. (Claude produced it for me 😎)

• Connect n8n to your Sheet. Read one row and send it by email. No AI yet, just the pipeline working.

Block 2 : Add Intelligence

• Create an Anthropic account, obtain your API key, and make your first manual API call from n8n. This is where the system starts to think.

• Build the prompt. This is the most important step: crafting the instruction that takes raw data from three functions and turns it into a clear, executive-readable paragraph.

• Connect everything. Sheet → n8n → Claude → email digest. Automate the trigger. Watch it run.

The result of this process is what you will see in the next edition.

If this article made you curious about agentic workflows, start here:

• n8n Documentation (n8n.io/docs): If you want to start building your own automation, the getting-started guides are genuinely accessible even without a technical background.

To keep things transparent under the EU AI Act: I used Claude (Anthropic) to help me structure this text, clean up my grammar, and generate the mock data for the tutorial. The ideas and final words are entirely my own.

See you soon. 👋

Disclaimer: The information provided in this newsletter and related resources is intended for informational and educational purposes only. It reflects both researched facts and my personal views. It does not constitute professional advice. Any actions taken based on the content of this newsletter are at the reader’s discretion.