Skip to content

How Ulern turns practice into evidence

The walkthrough, 1 minute 16 seconds, captioned. Real Ulern screens from a demo academy we keep for this; the people in it are invented.

A programme usually finds out what a participant learned at the end, with a test. Ulern finds out as they go: every answer in a session is evidence about a topic, and the topic’s standing moves only when the evidence supports it. This post walks through what that looks like, using the video above.

Everything on screen is the product, recorded in October 2026 in a demo academy we keep for this. The people are invented, and so are the participants’ answers (given by simulated participants). Every question, explanation and reply in it is the AI tutor’s own (on screen it is called Uli). For the shape of a session from start to finish, see what happens in an Ulern session.

Activities that ask the participant to do something

A session opens with a short check, and “I’m not sure” is always one of the answers. It is an honest answer, not a wrong one: it tells the AI tutor where to start. From there, the AI tutor explains an idea once and asks the participant to use it straight away, because doing beats re-reading.

The activities vary on purpose. In the video, from three participants’ first sessions, you see a scenario that sets an idea in a situation like the participant’s own, a straight multiple choice, a question with more than one right answer, statements to sort into what they are, and an answer the participant writes out in their own words, which the AI tutor reads and replies to. Ulern also uses fill-ins, ordering, matching, numeric answers and, where a subject calls for it, code the participant runs.

Every answer is evidence

Each answer lands on a learning target as evidence, and a target moves up through its bands (familiar, competent, mastered) only when the evidence says so. At the end of the session in the video, two topics read “Not started → Competent”. The participant’s plan shows each milestone’s topics as proven, shaky or not yet touched.

Not all evidence counts the same. Picking the right answer from a list says less than producing it, so a topic cannot reach competent on picked answers alone: it takes an answer the participant produced. Mastered takes one at the depth the topic aims at. Most tests never get that far.

Depth, not just right answers

Every topic aims at a level: understanding it, applying it, analysing with it, evaluating with it. The level is part of the plan, and the evidence has to reach it. A participant who can define a rate but cannot yet use one to judge a change is shown as exactly that.

What the reports show

The participant’s reports show where each topic stands, from familiar to mastered, and which topics are starting to fade, with a short review session to keep them. The supervisor sees the same evidence: for each topic, the level it aims at, the participant’s band, the confidence behind it and the answers it rests on.

The same record answers two questions. The participant sees what to do next. The supervisor sees what the participant can actually do, and how anyone knows.

For what a supervisor can check about the teaching itself, see how to check what Ulern teaches. To see it with your own subject, book a demo.

← All posts