Guided tutorial
The console carries a guided tutorial that teaches the platform from inside the platform. It highlights the real buttons on the real screens, waits while you click them, and leaves you holding a small working model you built yourself rather than a demo somebody showed you.
It is the fastest way to take a new colleague from "what is this" to a data set, a resource network, a measurement and an event, in about the time a presentation would have taken.
Starting it
Two entry points, both leading to the same flow:
- The user menu, top right. The entry marked Guided Tutorial, available on every page of the console.
- The home screen. The "New to DataHub?" bar, for anyone who has just signed in for the first time.
Either one opens Start guided tutorial?, which offers two ways in:
- From the beginning, the full tour.
- Or jump to a chapter, which plays from the chapter you pick through to the end. If the chapter you choose needs something an earlier one would have built, the tutorial creates that first rather than leaving you stuck.
Pick the industry it speaks in
Before the first step you choose a scenario: Oil & Gas (the default), Tech / IoT, Manufacturing or Energy.
The choice changes the vocabulary rather than the mechanics. The data set you create, the resources you connect and the measurement you add are all named for the industry you picked, so the model you finish with reads like a small piece of your own operation. That matters more than it sounds: the point of the tour is to recognise your own plant in it, and generic example names are what usually stop people doing that.
The six chapters
Each chapter opens with a short panel explaining why the thing exists before anything gets clicked, then walks you through building it.
| Chapter | What you do | What you are left with |
|---|---|---|
| Datasets | Create a data set, name it, give it an external id | The boundary everything else in the tour lands inside |
| Resources | Build a small network of resources and connect them | The graph, in miniature, with your own relationships on it |
| Timeseries | Add a measurement and attach it to a resource | A series that hangs off a thing rather than floating alone |
| Correlation | Look at your series alongside two siblings in Insights | The point that one signal is rarely the story |
| Events | Record what happened to the asset as events | Readings and happenings tied to the same network |
| Build a network | A larger network is built for you, highlighted and explained, then handed over for you to extend | A model you can keep and carry on with |
The order is deliberate and it is the same order the rest of this section is written in: a container, the things in it, what they measure, how those measurements relate, what happened to them, and then the whole shape at once.
How it behaves while it runs
- It waits for you. A step with an action does not advance until you actually perform it, clicking the highlighted button or typing a name of at least a few characters. Nothing is filled in on your behalf where the point of the step is to do it yourself.
- It blocks everything else. While a step is live, clicks outside the highlighted area are swallowed, so it is difficult to wander off mid-step and end up lost. Where a step deliberately hands you the canvas, the canvas becomes clickable again.
- It follows you across screens. Chapters run on the data sets, resources, time series, insights and events pages, and the tutorial navigates between them and picks up where it left off.
- Progress is a chapter strip at the top of each bubble, so you can see how much of the journey is behind you rather than counting steps.
- You can leave and come back. Every bubble carries Back, Quit and Next, and some carry Skip. If you quit or navigate away mid-tour, the next page you open offers a banner with Resume, Restart and Quit.
- It speaks the console's language. English and Norwegian Bokmål, following the language setting rather than a separate control.
It creates real data, and tidies up after itself
This is the part worth knowing before turning a team loose on a production tenant.
What the tutorial creates is real. A data set, resources, a time series with datapoints, a couple of events, and two extra series for the correlation chapter, all ordinary objects in your tenant, visible to anybody who can read that data set. Nothing is a simulation or a sandbox mode.
Three behaviours follow from that, and they are all deliberate:
- The last step asks. It offers Keep it, which leaves the network in place to inspect or carry on with, or Remove tutorial data, which deletes the data set, resources, series and events this run created.
- Starting again cleans up first. A fresh run deletes the previous run's objects before it begins, so repeated runs by the same person do not pile up.
- Each run gets its own identifiers. A run stamps a token so the things it creates get unique external ids rather than colliding with the last run's.
Deletion is best effort: if something cannot be removed, the tutorial carries on rather than stopping to complain. In a shared tenant it is worth a glance at the data set list afterwards.
A test tenant if you have one. If you do not, it is safe enough in a real one provided people know to finish with Remove tutorial data, and provided the tutorial data set is not left sitting in a list of production data sets long enough for somebody to mistake it for real.
What it needs
- Permission to create a data set, which is the all-data-sets grant rather than a grant on individual data sets, because a data set is the unit access is granted on. Without it the first chapter cannot complete. How permissions work →
- The feature enabled in the deployment. It is on by default; an operator can turn it off
everywhere with
tutorial.enabled=falsein the console's configuration, which also removes the menu entry and the home-screen bar.
When to reach for it, and when not
Reach for it on somebody's first day, for a team that has read about the platform and not touched it, and as the fastest way to answer "what does this actually look like" without booking a demonstration.
Read your first hour instead if you would rather work from written steps you can read ahead in, skim, or follow while building something of your own. It covers the same journey in the same order, which means the two can be mixed: take the tour, then use the written version when you build the real thing.
- Console tour: what each screen is for, read rather than walked
- Your first hour: the same journey in written form
- Data sets: the container the tutorial starts with
- The resource graph: the chapter that does the most work
- Time series and Events: what the model produces
- Insights: where the correlation chapter takes you