Sana's AI assistant is a powerful tool designed to help you create engaging and effective learning content and answer any questions you may have.
What can Sana’s Assistant do?
Content Creation: The AI assistant can generate paragraphs, lists, and summaries, making it easier for you to create comprehensive and informative courses. It can also generate completely original images in various themes such as 3D render, digital art, oil painting, and abstract.
Universal Search: The AI assistant is equipped with a universal search feature that can find information from all your company's Sana-created content as well as any integrated apps.
Easy Access: The AI assistant is accessible via phone, tablet, or computer, making learning convenient and flexible.
Continuous Learning and Improvement: The AI assistant is always learning and improving, with more abilities and uses becoming available as they are developed.
Use Cases for Sana's AI Assistant
Streamlining Content Creation: Sana's AI Assistant helps course authors easily generate comprehensive content, including paragraphs, lists, summaries, and even original images, reducing content creation time and effort
Enhancing Employee Onboarding: For efficient onboarding, HR managers can use Sana's AI Assistant's universal search feature to quickly find answers to questions generated from learning resources from the company's Sana-created content and integrated apps
Facilitating Continuous Learning: To foster a culture of continuous learning, organizations can use Sana's AI Assistant to provide personalized learning recommendations to each employee, keeping them engaged and focused on beneficial content.
Utilizing the assistant with Universal Search
Asking a question: Trigger the assistant by typing your prompt and pressing enter and that’s it! For example, your prompt might be “how do I report VAT for telephone expenses?“, once you press enter the assistant will interpret your query as a question, and give you a direct answer using the matching results as sources. Scroll down if you want to see the matching results that the assistant use for context
Conversing with the assistant: you can always follow up on anything the assistant says in the chat field, ask clarifying questions or rewrite the answer (e.g., “can you make that into 3 bullets?“)
No answer? If the assistant can’t find an answer based on the matching sources, it will tell you so. Sometimes it can help to rewrite your question in another form!
Utilizing the assistant: In Create
The assistant is always at your hand in the editor, simply create a new text block on a card and click the assistant button or use the generate button on cards such as the poll card:
The assistant button: in a new text block the assistant button option is presented
Below are the assistant options:
Assistant button options
Continue writing: AI continues where you left off
Summarize: AI summarizes all the text on the card prior to the text block
Brainstorm: AI brainstorms ideas for you using the information prior to the text block
Add an image: uses AI to select a stock image that is relevant to the content of the card
Generate an image: generates an AI image relevant to the content of the card
Generate from file: Check out our full guide here which shows you how you can generate a whole course from a PDF!
Typing /generate takeaway in a new paragraph the assistant reads the previously written text in the course to produce a takeaway
Generate button (on eg. question cards, polls etc) will read either cards/modules prior or after that card depending on the option selected
💡 Using the Continue writing option in conjunction with a prompt is like using the assistant in Universal Search. Eg. if you write, “Write a short snappy sentence capturing the essence of the above” and ask the assistant to continue writing, it will follow your lead.
Check out the prompt cheat sheet for some more ideas!
More info
Below is some more information about our interactive elements and the context used by AI to generate:
Image generation:
You can use it directly in your content with the assistant button (as seen in the screenshot above) or open any image modal (image blocks, cards, or course images) and select the “generate” option. Write your prompt, choose a style, then pick your favorite!
If you open the image modal again, you’ll find the historical images you generated before to use again.
Takeaway generation:
Type ‘/generate takeaways’ in a new paragraph in your content
The scope of the context AI uses is the text that is on the same card as the takeaway and precedes the takeaway (so any text that comes after the takeaway is not considered context).
Question generation:
Question card: The scope of the context AI uses is all cards that precede the question card within the same module
Question block within the Page card: Same as takeaways, the scope of the context is the text that is on the same card as the question block and precedes the question (so any text that comes after the question block is not considered as context)
Poll generation
There are three types of polls which you can generate dependent on placement within a module and what type of information you want to gather. For example, an introductory poll would offer some insight into how learners feel about something prior to engaging with the respective content.
Check-in poll: The scope of the context AI uses is all cards that precede and succeed the poll card within the same module.
Introductory poll: The scope of the context AI uses is all cards that succeed the poll card within the same module.
Reflective poll: The scope of the context AI uses is all cards that precede the poll card within the same module
Assessment generation:
The scope of the context AI uses is, at this point, always up until the previous question. So in this case, the assessment would only look at the card before. If you would remove the questions that come in between the cards, you'd be able to generate questions based on all the content. One trick could be to temporarily move the questions into a separate folder while generating assessment questions, and then move them to the right position. On a side note, we will be improving how easy it is to understand which context is being used for the question generation.
Related articles: