No special permission is required to use Stella — it is available by default as a tab on the navigation bar, and what it shows you follows your existing role on the platform.
- Learner: can ask Stella about their own progress and get redirected to training.
- Manager: can ask Stella about data covering their team.
- Administrator: can ask Stella about platform-wide data, and is the only role that can remove the Stella chat from the navigation bar.
Contact your Administrator if Stella is not visible on your homepage.
Stella is a conversational AI chat, accessible from a dedicated tab on the navigation bar. A user types a question in plain language, and Stella responds in one of two ways depending on what is asked:
- 1Data question — Stella answers directly in the chat with a metric, table, or chart, scoped to the user's role and downloadable as a CSV file or image.
- 2Content question — Stella identifies a matching training or content item on the platform and redirects the user to it.
Stella does not explain concepts or teach content, and it never answers using knowledge outside the platform. If it cannot find a data answer or a matching content item, it says so rather than guessing. Sections IV and V of this article give prompting tips and a ready-to-use prompt library for every role.
- Stella chat: the conversational AI chat, accessible from a dedicated tab on the user navigation bar.
- Prompt: the question typed into Stella. Its phrasing determines whether Stella returns data or a redirect.
- Follow-up: a question asked in the same conversation that refines the previous answer (for example, excluding suspended users).
- Data scope: the set of records a user can query, determined by their role (learner, manager, administrator).
- Export: the option to download a data answer as a CSV file (tables) or an image (charts).
- Redirect: the action of pointing a user to a training, course step, or catalog item that matches their question.
- Out-of-scope response: the short, neutral message shown when a question falls entirely outside platform content.
I — Ask Stella a question
- From the navigation bar, open the Stella tab.
- Type a question in plain language.
- Stella determines whether the question is about training data or about finding content, and responds accordingly (see Sections II and III).
Result: The user gets an answer or a redirect without needing to browse the catalog or build a report manually.
II — Data and analytics answers
When a question is about training data, Stella can return:
- A single metric (for example, a completion rate).
- A table of results, downloadable as a CSV file.
- A chart, downloadable as an image.
The data returned is always scoped to the role of the person asking:
| Role | Data visible in Stella | Typical questions |
|---|---|---|
| Learner | Their own progress only. | Progress, scores, deadlines, certificates, completed courses, and finding training. |
| Manager | Data for their team. | Completion and progress of team members on a training, who is late, who has not started. |
| Administrator | Data for the whole platform. | Platform-wide users, groups, courses, completion rates, connections, and charts. |
III — Content redirects
When a question is about finding training, Stella:
- Looks for a related training or content item available on the platform.
- If a match is found, redirects the user to it.
- If no match is found, gives a minimal contextual response and states that it cannot redirect.
IV — Write better prompts
The way a question is written directly affects the answer. These five habits turn a vague question into a precise, reusable answer.
FindIs thereWhat training can help return a redirect.
this monthlast monththe past 6 months
by user in a tableas a bar chartas a pie chartas a line chart
Can you exclude the suspended users?How many didn't start the training?
From vague to precise
Result: A precise answer on the first try, ready to reuse: tables download as CSV files and charts as images.
V — Prompt library
Ready-to-use prompts, grouped by role. Copy a prompt into the Stella tab and replace anything shown as [training name] with the exact name used on your platform. Dashed rows are follow-ups: ask them in the same conversation, right after the prompt above.
Learner — Track your progress
Learner — Find training and get around
Content questions redirect you to a training, course step, or catalog item. If nothing matches, Stella states that it cannot redirect you.
Manager & Administrator — Follow training completion
Managers can use these prompts too: their results are limited to their own team.
Administrator — Platform, catalog, and activity
Ask for a chart type explicitly (bar, pie, line). Charts download as images and tables as CSV files.
VI — Manage Stella
Only an Administrator can remove the Stella chat from the platform. Learners cannot remove or hide it from their own homepage.
- An administrator removes the Stella chat platform-wide from AI & Adaptive features.
- From the same menu, administrators can also customise Stella's name at the bottom of the page.
Platform usage of Stella is tracked through the number of questions asked, the redirect rate (questions that resulted in a redirect versus total questions), and the removal status.
VII — Scope and limitations
Platform-grounded answers only
Stella only generates responses based on content available on the platform: course content, the catalog, and learner data. It does not generate answers from its own general knowledge. If a question has no match — whether a data answer or a content redirect — Stella explicitly says it cannot answer rather than guessing.
General questions
If a user asks a general question not directly linked to a specific training, Stella gives a minimal answer and redirects to relevant platform content — always a training, course step, or catalog item. If no relevant content exists to redirect to, Stella only states that it cannot answer.
Out-of-scope questions
If a question is entirely outside the platform context, Stella explicitly says the question is out of scope. This message is short, neutral, and non-judgmental. For example:
- "What's the weather tomorrow?" — out of scope.
- "Who won the World Cup?" — out of scope.
- "Can you teach me how to negotiate?" — Stella points to a related training if one exists; otherwise it says it cannot help.
- Stella never generates content that is not grounded in platform data.
- Stella never acts as a tutor, coach, or general chatbot.
- Stella never answers a question using external knowledge, even if platform content is insufficient.
FAQ & Troubleshooting
-
Issue: Stella is not returning the expected data for the user.
Solution: Stella scopes results to the user's role. A learner will only ever see their own progress, a manager only their team, and an administrator the whole platform. This is expected behaviour and not an error. If the result still looks wrong, send feedback with the thumbs up / down.
Issue: Stella keeps saying it cannot answer.
Solution: Stella only answers from data and content that already exist on the platform. If no data matches the question, or no training, course step, or catalog item matches it, Stella will say it cannot answer rather than invent one. Confirm the relevant data or content is available, then rephrase using the exact name of the training, group, or user (see Section IV).
Issue: A user got a data answer when they expected to be redirected to training, or the other way around.
Solution: Stella decides how to respond based on how the question is phrased. Ask the user to rephrase — for example, asking to "find a course on X" rather than "how am I doing on X" — to get a redirect instead of a data answer.
Issue: The answer is returned as a number when a table or chart was needed.
Solution: State the format in the prompt, for example "by user in a table" or "as a bar chart", or ask for it as a follow-up in the same conversation.
Issue: A user expects Stella to explain a concept and it does not.
Solution: This is expected. Stella redirects to content or answers data questions — it does not teach or explain content itself. Open the training it suggests.
Issue: Stella's conversation history does not load.
Solution: This is a known potential issue when loading conversation history. Ask the user to refresh the page and try again; if the issue persists, contact support.
Issue: Stella is not visible on the homepage.
Solution: Contact your Administrator. Only administrators can remove or restore Stella, from AI & Adaptive features.
-
Is Stella the same as Forge?
— No. Stella is a separate feature. It does not teach or explain content; it answers data questions and redirects learners to existing content.
Where can I find example prompts?
— Section V of this article contains a prompt library organised by role: learner, manager, and administrator.
Do I need to start a new conversation for each question?
— No. Follow-up questions in the same conversation build on the previous answer, for example "Can you exclude the suspended users?"
Can a learner remove the Stella chat from the homepage?
— No. Only an Administrator can remove the Stella chat, from AI & Adaptive features. Learners cannot remove or hide it themselves.
What happens if Stella cannot identify relevant content?
— It provides a minimal contextual response and clearly states that it cannot redirect the learner to relevant content.
Can Stella's data answers be shared or reused outside the platform?
— Yes. Tables can be downloaded as CSV files and charts as images.
Does Stella let a manager see other managers' teams?
— No. A manager's results in Stella are limited to their own team. -
Contact Support