Find Answers and Training with Saga - Article

Open a conversation

Where enabled, select the assistant control in the platform header to open the chat alongside your current page. This lets you keep your place while asking a question. The sidebar can offer recent conversations and a way to start a New chat. Use Open full chat when you need more space or your full personal chat history.

A navigation-menu entry can also open the full assistant page where the platform shows that link. The name can be customized. If the menu link is absent but the header assistant remains, that can be an intentional navigation choice rather than a feature failure.

1. Describe what you are trying to understand or achieve. Add useful context without unnecessary personal or confidential information.
2. Send one clear question. While the assistant is working, wait for the response before submitting the next message.
3. Read the answer, inspect the sources where provided and open the relevant original material.
4. Ask a follow-up to narrow the explanation or compare alternatives. Selecting a suggested follow-up sends that question; it is not merely a filter.

Ask for a useful outcome

Read sources, not only the generated answer

Where source references are supplied, use the inline reference or the Sources control to inspect them. A source can lead to the relevant course module, page, file, activity description or video position. Video navigation depends on the indexed text and reference information available for that recording.

Open the original source to confirm the wording, date and context. A citation is useful navigation, not proof that every generated interpretation is correct. If the response has no usable source, ask which available material supports it and verify important information independently.

For example, an answer about shutting equipment down may summarize general course advice. Before applying it to safety-critical machinery, read the organization’s current approved procedure. Learning guidance is not permission to disregard operating or safety requirements.

Discover related training, including paid offers

A useful answer can open a further learning opportunity. You may ask how to achieve something, receive an explanation from accessible material and discover another training that develops the topic or provides the practical skills you need next. You do not have to know that training’s title in advance. Saga can recommend eligible Storefront offers from their activity descriptions and keywords without revealing lessons you are not permitted to access.

Recommendations can include free training and, where commerce is used, training with a price. A fee does not by itself make an otherwise eligible offer unavailable for recommendation. For learners, this connects an immediate question with a relevant next step. For organizations and training providers, it helps people discover related offers across the catalog and supports cross-selling or upselling when a paid course, workshop or more advanced program fits the learning need. The practical value applies equally to free training: useful material can reach people who would not otherwise know to look for it.

For example, a service partner who asks how to identify avoidable energy use may receive an introductory explanation and a suggestion to explore a practical energy-audit workshop. If the workshop is a paid offer, it can still be a relevant next step. This is an illustrative learning progression, not a promise that a particular training will be returned. Open the recommended offer to decide whether its outcomes, level and terms suit your situation.

Move from a recommendation to participation

Before enrollment, Saga can suggest eligible offers from the personal Storefront descriptions available to it. Open the training description to check the audience, delivery, schedule, price and participation conditions. Then follow the actual self-enrollment, approval or purchase process. A recommendation does not remove a fee, satisfy prerequisites or grant access to protected lessons before enrollment.

A description can be unavailable to the general Storefront discovery tool when self-enrollment is disabled. During an enrolled training, the activity-specific assistant can still have basic activity and module overview context. This difference does not mean it can reveal all protected lessons or that the training has become self-enrollable.

The assistant can use relevant profile and learning-history context to make an answer more appropriate. A recommendation is still not an assignment, a reserved seat, a guarantee of suitability or an automatic learning plan. Compare it with your objectives and any organizational requirements.

Help relevant training be discovered

For content owners, a clear activity description is part of making the offer discoverable—not merely promotional text. Explain the problem the training addresses, who it is for, its level, any expected prior knowledge and what participants will be able to do afterwards. Use accurate activity keywords that reflect how customers, partners and other learners describe the need. These descriptions and keywords give the assistant meaningful context for connecting a question with an appropriate offer; vague titles or generic claims provide less useful information.

For example, an illustrative description could say: “For service partners and facility coordinators who understand basic energy use. Learn to carry out a simple site walk-through, identify avoidable consumption and prepare a prioritized improvement plan.” This explains the intended audience, starting point and practical outcome more clearly than “Advanced energy workshop.” Keep the offer’s description, schedule, delivery and price accurate, and check that its discovery content is indexed and available to the intended audience. Better descriptions support relevance; they do not guarantee a recommendation or override access rules. For indexing controls, see Control AI Content Indexing and Protect Assessments.

Manage your own conversations

Use a new chat when changing to an unrelated subject. Earlier messages in a conversation provide context, so continuing a long unrelated discussion can produce less useful answers. Use the full-page history to reopen your own previous chats and the available delete action when a conversation is no longer needed.

Chats are personal to the user; another person’s direct chat URL is not a way to open their conversation. Deleting a chat affects the platform conversation record. Do not interpret that action as a guarantee that all provider or operational processing logs have been erased; see Understand AI Providers, Data Processing, and Governance.

The widget limits the length of an individual message. If a question is too long, split it into a clear initial request and focused follow-ups instead of pasting a complete document. Use the response-copy control when you need to reuse wording, but retain any source and accuracy checks.

Use the right context for training support

If you are already working inside a Course module or Learning Path, use the dedicated training assistant for questions about that activity. Its conversation is separate from general chat and follows the training’s accessible modules. See Use Saga During Training.

Troubleshooting