With Generate from file, we are making it easier and faster than ever to create engaging courses in Sana based on existing learning materials. Upload your files and watch Sana generate a course based on the text in the document.
How it works
When generating a course from the supplied PDF, Sana's assistant will go through the pages one by one and generate a corresponding card in Sana. The assistant will try to output text similar to the original page, but since it is generating text it will not be a one-to-one representation of the PDF. You can configure your course according to the desired length, level of interactivity and tone of voice (see below). It's important to note that the quality and structure of the generated course can vary depending on the complexity and formatting of the original PDF. While our tool does its best to create an organized course, it may require manual adjustments to achieve the desired outcome.
The existing feature version is most proficient at processing text-heavy PDFs, but we are actively progressing towards accommodating PDFs which include a variety of formats. We are also working on being able to extract text from images, which is common for presentations that have been converted into PDFs.
Configure your course
With the Generate from file feature, you have the flexibility to configure your course to fit your desired output. You can control various aspects of the course including course length, level of interactivity and tone of voice.
Course length: Choose whether you want a summarized version or a full-length course that aims to match the original text.
Interactivity: Adjust the level of interactivity, ranging from no interactive elements to highly interactive.
Tone of voice: Set the tone of voice to align with your desired communication style.
Interactive elements
If interactivity is configured, the assistant will automatically insert interactive elements, such as reflection, poll and question cards in the generated course to create an engaging course.
Additional info:
Images that are extracted in the generation can be moved and re-sized to your liking after inserting the cards into the editor.
The assistant extracts a title for each card based on headings in the PDF. The first card will always be a Title card with a generated title.
The assistant tries to read tables, but the output may vary depending on how the table in the original PDF is formatted.
If your PDF is too large to be processed, try splitting it into sections and upload them separately.
Instructions
Step 1: Upload file
In the editor:
Click the arrow next to the Add card button in the bottom left, then click Generate from file
Drag and drop your PDF or click Browse files to choose your file and the assistant will begin generating the course
Configure your course to suit your desired course length, level of interactivity and tone of voice
See a demonstration below of a course being generated from a PDF presentation about machine learning
💡 As content generation takes place, you are able to follow it in real-time. If you wish to halt the process, simply click on the pause button located in the lower left corner.
Re-generate cards
After the generation is done, it is possible to either re-generate specific cards or all cards. You can also save specific cards by clicking the 3-dot menu. These cards will then not be re-generated if you want to re-generate the rest of the course.
Step 2: Insert into course
If you are happy with the results, simply click Insert cards and and the generated cards are saved to your deck. From here, you can continue editing the course and add more interactive elements, customize it or publish it directly.
Best practices for generating courses from file
When using the "Generate from file" feature in Sana, here are some best practices to ensure you create effective and engaging courses:
Use text-heavy PDFs: The feature is most proficient with text-heavy PDFs. While the feature is being developed to handle a variety of formats, including extracting text from images, currently, text-heavy PDFs yield the best results.
Keep content on a single page: Since the system processes content one page at a time and may not interpret information that flows between pages accurately, ensure each point or section is confined to a single page for clarity.
Use high-quality images: While text should be the primary focus, any images included in the PDF should be high-quality and relevant to the content. This enhances visual appeal without overwhelming the system.
Utilize the course configurators: Use the course configurators to fine-tune your content and control how you want the output to be.
Review and edit: Be prepared to make minor manual adjustments. After generating the course, thoroughly review the content, re-generating specific cards as needed, and save those you are satisfied with.