What would you like to learn?

Search for the software, or for what you want to be able to do.

Common starting points:

Alongrail Learning · free courses

Learn software well enough to use it at work, by building a real project in it.

You choose a free course, open the software and share your screen. Alongrail shows you the next step from what’s actually there, while you do every click. In the second part of each session, you build it again on different data, on your own, and ask only when you need to. You finish with a project you built yourself, and you know you can do it again without help.

See the free courses →How a course works

28 free courses, including Excel, Power BI, SQL, Python and Figma. No payment and no card. Each course says which computer it needs. You choose what to share, you can stop at any time, and Alongrail never clicks for you.

Executive summaryRegional breakdownYour own question

Live now — enrol free and start today

Power BI Foundations

Build a professional sales dashboard from raw business data.

About 12 hours·Windows

What would you like to learn?

Learn software well enough to use it at work, by building a real project in it.

Live now — enrol free and start today

Build a professional sales dashboard from raw business data.

A sales summary workbook someone else can open and read, where every figure traces back to the data.

Read a database you did not build: find rows, count and total them, group them, and bring two tables together.

Build and publish an interactive dashboard anyone can open, from three raw exports.

Deeper modelling and calculation, where the right answer isn’t obvious.

Build a report someone can open at a link, from three connected files.

Answer a manager’s questions from a file you did not make, in a notebook you keep.

Build a base your colleagues can use: two linked tables, the views that answer questions, and a share.

Turn a raw export into a service desk anyone can read, published at a link.

One messy export becomes a trustworthy table, a summary by group, a chart and a saved results file, as a script in base R.

Put the work in, make it true, and get the answer out: a board, one sprint, and the manager’s questions answered on a dashboard.

A project of yours, on GitHub, with a history you can defend.

Turn a raw draft into a report somebody else can read, navigate and trust.

Turn two raw exports into a CRM anybody can trust.

Turn an export into a sales pipeline you can read.

Turn a raw export into a summary someone else can open, read and trust.

One page, written by hand from a client’s content pack, with the browser read back at every step.

Build the deck somebody else asked for, in the browser, and check it before you send it.

A brand kit anyone could work from, and a campaign in three formats that obeys it, on a free account.

Two live automations from a form to a spreadsheet and an inbox, their runs read and their cost counted.

Somebody else’s folder made a workspace a colleague could open: extensions, settings and formatting the project decides, and a history.

A Search campaign built from a written plan and left as a draft an owner could approve, in an account that holds no money.

A Facebook and Instagram campaign built from a written plan and left as a draft an owner could approve, with no payment method on the account.

A research briefing pack made from eight supplied documents — titled notes, a study guide, a data table, a slide deck and an audio briefing — shared by a link.

A one-page website for a small firm, published at your own address: the AI drafts the structure, you read every change before keeping it, and you type every fact yourself.

The three main screens of a booking app, built from a first draft with components and styles, and shared as a prototype anyone with the link can click through.

A collection that tests a shop’s order API: requests with parameters, a body and headers, two environments with the key kept out of the files, checks that fail and pass when they should, and the whole collection run at once.

A six-page report from plain text: headings that number themselves, figures, a table built from a CSV file, a bibliography, a contents page, a layout kept as a template, and the PDF exported.

An hour of your time, one step at a time.

Build it with guidance. Then build it again yourself.

Coming soon

The final level: building and defending analysis someone else will rely on.

A monthly report pack built from more than one file: lookups, PivotTables and the checks that catch them out.

A model other people depend on, and the audit trail for it.

Answer questions one pass cannot: outer joins, subqueries, set operations, CTEs, and changing data safely.

Window functions, recursive queries, making queries fast on purpose, and designing a schema others will query.

Level-of-detail expressions and real data preparation: the numbers that hold when the question changes.

Take someone else’s published workbook that is wrong, slow or unreadable, and fix it.

One report over several sources, that survives being changed.

A report other people depend on: team workspaces, scheduled delivery, governance.