Courses

Courses are the core training content in Aimee.ai. Each course defines an AI-powered conversational simulation that learners will complete.

Accessible to: Trainers, Workspace Owners, Admins

Course list

Navigate to Content Library in the sidebar to see all courses in the workspace on the Courses tab. For the full lifecycle view (drafts, versions, duplicating, templates, purchased courses), see Course Management.

Each course row shows:

• Title and description
• Status - Draft or Published
• Version label (e.g., v1.0)
• Last updated date

Use the search bar to filter courses by title, the status filter to show only drafts or published courses, and the tag filter bar to narrow by topic.

Creating a course

1. Open the Courses tab in the Content Library.
2. Enter a title in the inline create form.
3. Click Create Course.
4. Choose a starting point:
   • Blank course - open an empty editor and build the content yourself.
   • Generate with AI - answer a few questions and Aimee drafts slides, questions, and narration for you.

The new course opens in the Course Editor as a draft.

When you choose Generate with AI, one of the steps asks how you want to brief Aimee. Alongside defining what's in and out of scope, or pasting freeform notes, you can choose Upload reference files to hand Aimee your own source material - a policy document, a slide deck, or a procedure. Upload up to five files (PDF, Word, PowerPoint, plain text, or Markdown; 25 MB in total). Aimee reads them and shows you a short, editable summary you can adjust before generating, so the course is grounded in your material. If your workspace owner has turned this option off, it won't appear.

Keeping the files you upload

If your workspace has the Knowledge Base turned on, the upload step also offers Also keep these files in your knowledge base, ticked by default. Leave it ticked and the documents you just uploaded stay in your Knowledge library rather than being discarded once the summary is written - so the next course, scenario, or question to Ask Aimee can use them without you uploading them a second time. They count towards your workspace's storage allowance like any other knowledge document.

Each file tells you what happened to it next to its name: Added to your knowledge base, Already in your knowledge base (you'd uploaded the same file before, so nothing is duplicated), or the reason it couldn't be added - for example if you've reached your storage limit. Your summary is produced either way; a problem keeping the file never costs you the summary.

Untick the box and the files behave as they always have: read once, then discarded. If your workspace doesn't use the Knowledge Base, the option doesn't appear and files are always discarded.

Generate from your Knowledge library instead

If the material already lives in your Knowledge library, you can ground the course on it directly instead of uploading it again - Aimee retrieves the relevant passages, cites the source on each slide, and can keep the course up to date as the source changes. See Building a Course from Knowledge.

Start from a template

If you've already saved reusable templates, open the Templates tab of the Content Library, open the template, and click Save as new course - see Course Management.

Course Editor

Click a course row to open the Course Editor. Every per-course setting is grouped into a row of tabs at the top of the page:

Tab	What it's for
Overview	Course title, description, tags, and default narration voice - the at-a-glance card.
Settings	Learner brief, scoring settings, and the underlying runtime configuration the AI uses - see Basic information, Learner Brief, Runtime Configuration, and Scoring settings below.
Content	Build the slides, scenarios, quizzes, matching, acknowledgement, resources, and other blocks. Most of your authoring time lives here - see Course Content & Voice Narration.
Constraints	Pre-publish gates the course must pass before it can be published - for example, "course must contain at least one block" or "every quiz block has a correct answer marked". The tab lists what's still outstanding so you know exactly what to fix.
Playback	Controls how the learner experiences the course at launch - autoplay defaults, narration locks, and whether the workspace's Branding Intro is included before this course. See Playback below.
Questions	Configures the Knowledge Check that runs at the end of the course - number of questions, difficulty, and how the AI samples from the course content.
Marketplace	List the course on the Marketplace, set its price, or unpublish it.

A Recovery tab also appears here, but only when an AI generation or import has failed and a block needs your attention to retry or repair. On a healthy course it stays hidden.

The day-to-day flow lives in Content and Settings. The other tabs come into play at publish time or when something needs course-correcting.

Basic information

• Title - the course name
• Subtitle - an optional one-line tagline (up to 80 characters) shown under the title on the Library, Catalog, Learning Path, and Marketplace surfaces. Use it to communicate the audience or value of the course at a glance. Leave blank if you don't need one.
• Description - internal description (not shown to learners)
• Default Narration Voice - the default voice used for AI narration across this course's content blocks

Course Content & Voice Narration

The Course Content tab lets you build a structured set of blocks the learner progresses through before or during the simulation. Blocks are grouped into modules (chapters) so larger courses scan as a clear outline rather than a long flat list - see Modules below for how to create, rename, and reorder them.

Use the Import Content menu to pull a whole file in - each page, slide, section, or heading becomes its own block:

• PDF - Aimee reads the PDF with AI and writes one editable slide per logical section, lifting embedded images out into the workspace image library so you can swap or re-use them. You see a cost preview before the import runs.
• Markdown (.md / .markdown) - Aimee reads the Markdown with AI and produces one editable slide per logical topic. # and ## headings are a strong default boundary, but tightly related sub-headings can be merged into a single slide. Inline images embedded in the file (data: URIs) are pulled into the workspace image library; remote ![alt](https://…) images are kept as direct links. Code blocks, tables, and lists carry through as semantic HTML. You see a cost preview before the import runs.
• PowerPoint (.pptx) - one block per slide. Speaker notes become voice narration for that slide. Side-by-side (two-column) slide layouts are kept rather than being flattened into a single column, and any charts in the deck are recovered as editable charts - either as a standalone Chart block, or inline within the slide when the chart shares the slide with other content.
• PowerPoint (AI) - Aimee reads the same .pptx file with AI and recognises richer source structure: a numbered question-and-options slide becomes a quiz block, an attestation slide becomes an acknowledgement block, everything else lands as an editable slide. The AI may merge tightly-related adjacent slides into one course slide and split content-dense slides into several. Speaker notes carry through to voice narration on the resulting slides. Embedded images carry through to the workspace image library. You see a cost preview before the import runs.
• Word (.docx) - one block per Heading 1 (falls back to Heading 2 if the document has no H1, or becomes a single slide if it has neither). Lists, tables, bold / italic, and embedded images all carry through.
• Word (AI) - Aimee reads the same .docx file with AI and recognises richer source structure: a numbered question-and-options section becomes a quiz block (with the correct answer pre-marked), an "I have read and understood…" section becomes an acknowledgement block, and everything else lands as an editable slide. Embedded images carry through to the workspace image library. You see a cost preview before the import runs. When the AI is unsure, it falls back to a plain slide so source content is never lost.

Each import creates a new module named after the file you imported, and drops its slides into that module - so a course built from onboarding.pdf and safety.pptx ends up with two clearly named chapters.

If your course already has content, Aimee asks whether to append the new file at the end or replace the previously imported content (manually added media, quizzes, and other blocks are preserved when you replace).

Or add blocks by hand from the Add Block menu:

• Accordion - a collapsible "click each row to expand" block - useful for FAQs, step-by-step instructions, or layered explanations the learner can reveal at their own pace.
• Acknowledgement - ask the learner to type their name to confirm they have read a policy.
• Boxed List - a read-only list of items, each shown in its own coloured box, with no order or sequence implied. Use it for a set of responsibilities, principles, or "things to remember" that read as peers rather than a numbered sequence.
• Chart - a pie, bar, or line chart over a small set of values - useful for proportions, comparisons, or trends. PowerPoint charts you import are turned into editable chart blocks automatically.
• Feedback - ask the learner for free-text feedback at a specific point in the course (after a module, on completion, etc.).
• Flip Cards - a grid of cards the learner clicks to flip over and reveal the back. Use it for term/definition pairs, prompts and answers, or short reveals.
• Matching - click-to-place exercise: pair terms with definitions, order steps, or sort into categories.
• Media - embed a video or audio clip (YouTube, uploaded video, image).
• Quiz - knowledge check with one or more questions.
• Resource - downloadable take-away files (policy PDFs, job aids, templates, checklists).
• Scenario - open-ended AI-driven roleplay the learner works through in conversation.
• Slide - rich-text slide with optional images and voice narration. The rich-text toolbar also includes a code snippet button for inserting a syntax-highlighted code block with a language tag and copy-to-clipboard affordance - handy for technical training. Use the Size dropdown (next to the paragraph/heading style) to make selected text Small, Large, or X-Large - it works on any run of text, including a few words inside a bullet list, so you can emphasise a phrase without turning it into a heading; Normal clears the size. When the cursor is inside a list, a list-marker button appears in the toolbar so you can change that list's bullets or numbering: bulleted lists can use Disc, Circle, Square, Triangle, Arrow, or Check markers, and numbered lists can use Decimal (1.), Lower alpha (a.), or Lower roman (i.). The style applies to the whole list, so one list can use check-marks while another on the same slide uses arrows. The size, along with any text colour or highlight you apply, shows the same way to learners. Slides can use a one-column layout (text only, today's default) or a two-column layout that puts the text on the left and a single image from the workspace image library on the right; pick the Image width (20% / 30% / 40% / 50% / 60%) so the image gets as much room as it needs without crowding the text. On phones and other narrow screens the image stacks above the text so the slide stays legible. Voice narration on a two-column slide is generated from the left-column text only - the image is silent. Save is disabled until you've picked an image.

When the cursor is inside a list on a slide, the list-marker button opens a picker for that list's bullet or numbering style:

Templates lose the two-column flag

Two-column slides round-trip through duplicate-course, but the template export → re-ingest path currently flattens them back to one-column. If you regenerate a template, recreate the two-column layout in the new course.

Browse uploaded files

Click Browse Files in the Content tab toolbar to see every file uploaded into this course - PDFs, PowerPoints, Word documents, Markdown files, images, and the audio files generated for voice narration. You can download or delete any file, and choose which PDF should be treated as the course's source.

Duplicate a block

Each block has a copy icon beside the trash icon. Click it to drop an independent copy of that block immediately after the original - handy when you want a quiz, slide, or scenario with the same shape but different wording. Media, voice audio, and (for resource blocks) attached files are all copied; editing the duplicate never changes the original.

Moving between blocks while editing

Every block editor has a Prev / Next control in its header, with a counter showing where you are in the current module (for example, Block 3 of 6) and the module's name underneath. Use it to work straight through a module without returning to the Content tab between blocks.

Next on a module's last block carries you into the first block of the next module, and the module name in the header changes so you can see you've crossed a chapter boundary. On the very first and very last block of the course, the matching arrow is greyed out.

If you have unsaved edits when you move, Aimee asks what to do:

• Save & continue - saves this block, then opens the next one.
• Discard - moves on without saving. If the block was one you had just added and never saved, it's removed rather than left behind empty.
• Keep editing - stays where you are.

When a block isn't complete enough to save - a slide with no title, an accordion with only one section - Save & continue is unavailable and the prompt tells you what's missing. Discard and Keep editing still work, so you're never stuck on a block you'd rather abandon.

The same prompt appears if you click Back, refresh, or close the tab with unsaved edits.

Modules

Modules are the chapters of your course. Each block belongs to exactly one module. The Content tab lays modules out as a board - each module is a column, side by side, with its blocks stacked inside it - so you can see several chapters at once and drag blocks between them instead of scrolling one long list. When there are more columns than fit on screen, the board scrolls sideways. Every course starts with a single default module called Course content - you can rename it like any other module, but it can't be deleted, so a course always has somewhere for its blocks to live.

Authoring. From the Content tab:

• Click + Module to add a new column. Click a module title to rename it inline.
• Drag a module's grip handle sideways to reorder the columns.
• Drag a block by its grip handle into another column to move it across chapters; drop on a module header to insert it at the top of that module.
• Use the Add to dropdown next to Add Block to choose which module a new block lands in, so it appears in the right column straight away instead of at the end of the course. (It appears once a course has more than one module.)
• Open a module's kebab menu and choose Delete module to remove a module - its blocks are moved to the end of the Course content module so nothing is lost.
• An empty module is allowed, so you can lay out a chapter skeleton before importing the content for it.
• To move several blocks at once, tick the checkbox on each block (or Select all), then click Move to in the toolbar that appears and pick the destination module - the selected blocks land at the end of that module, keeping their order. You can also drag any one of the selected blocks and the whole selection moves together.
• On a narrow screen the board falls back to a single stacked column so it stays easy to use.

What the learner sees. When learners open the Contents panel from the top-left of the course player, lessons are grouped under their module names as numbered chapters, with a completion tick beside each lesson they've finished.

Resource Blocks

A resource block lets you attach one or more downloadable files alongside the lesson - policy PDFs, SOPs, spreadsheets, job aids, templates, checklists. Each file in the block gets its own label shown to the learner (e.g. "Policy PDF", "Signing template", "Onboarding checklist"), so related artefacts can live together in one block instead of a separate block per file.

Authoring. Click Add Block and choose Resource, then:

1. (Optional) Give the block a title and description - context shown above the file list.
2. Click Add files to pick one or more files. You can add up to 20 files per block.
3. Set a label for each file (optional; defaults to the filename).
4. Remove a file with the trash icon; replace the entire set by removing and re-adding.

File types accepted: PDF, Word, Excel, PowerPoint, CSV, TXT, ZIP. Uploads that exceed the workspace file-size cap are rejected, and file types outside the allowlist are refused at save time - not just in the picker.

What the learner sees. Each file renders as its own download card: a file-type icon, the label (or filename as a fallback), the filename and size, and a Download button that triggers the browser's save-as dialog using the original filename.

Resources are optional

Resource blocks are informational, not assessments - downloading is not a prerequisite for course completion. Use them for reference material that learners may want to keep after the course.

ZIPs are trusted at upload time

Files are stored as-is; there's no virus scan. Only upload materials you have authored or vetted.

Matching Blocks

Matching blocks are interactive click-to-place activities for assessment shapes a multiple-choice quiz can't capture naturally. The learner clicks an item in the bank, then clicks the target slot where it belongs. One block, three modes:

• Pair terms with definitions - e.g. match each ISO 27001 control to the risk it mitigates.
• Order steps - e.g. put the incident-response steps in the correct sequence.
• Sort into categories - e.g. sort each attribute into PII / Sensitive PII / Non-PII.

Authoring. Click Add Block and choose Matching, then:

1. Pick the mode (pairs, ordering, or categories).
2. Fill in a title and optional instructions shown above the activity.
3. Add at least two items:
4. For pairs, each item has a left (term) and right (definition).
5. For ordering, each item has a label; the order you enter is the correct order.
6. For categories, add at least two categories first, then tag each item with its correct category.
7. Optionally tick Allow partial credit and set a pass fraction (e.g. 0.75 means 75% of items must be in the right place).

Labels typed in one mode carry over if you switch modes, so you can prototype the wording and pick the shape that fits best.

Not for high-stakes assessment

Matching blocks are scored in the learner's browser (same as Knowledge Check quizzes on a slide). A motivated learner can inspect the page to see the correct answer before submitting. This is fine for practice-style inline assessment; for certification-grade testing, use a dedicated assessment workflow.

Accordion Blocks

An accordion block stacks two to eight collapsible sections the learner expands at their own pace. Use it for reference content the learner browses rather than reads top-to-bottom - definitions, pillars, FAQ-style lists, or "here are the five things you need to know" layouts.

Authoring. Click Add Block and choose Accordion, then:

1. Give the block a title (required) and an optional intro rendered above the sections.
2. Add at least two sections. Each section has a short heading and a rich-text body (lists, links, bold/italic, code snippets, embedded images).
3. Drag the grip handle on a section to reorder; click the trash icon to remove. You can have up to eight sections per block.
4. Choose a gating option:
   • None - the learner can advance whenever they like.
   • All opened - the learner must expand every section at least once before the course will let them move on.

What the learner sees. Each section starts collapsed with only its heading visible; clicking expands it to reveal the body. Opening one section doesn't close the others, so the learner can compare two sections side by side.

Flip Card Blocks

A flip card block is a grid of two to six cards the learner clicks to reveal the back. Use it for term-and-definition pairs, prompts and answers, or any short front/back reveal you'd otherwise put on a paper flashcard.

Authoring. Click Add Block and choose Flip Cards, then:

1. Give the block a title (required) and an optional intro.
2. Add at least two cards. Each card has a short front title and a rich-text back body, plus an optional image from the workspace image library shown on the front.
3. Drag to reorder; click the trash icon to remove. You can have up to six cards per block.
4. Choose a gating option:
   • None - the learner can advance freely.
   • All flipped - the learner must flip every card at least once before continuing.

What the learner sees. The cards lay out in a responsive grid - one column on a phone, two or three on a laptop. Clicking a card flips it over; clicking again flips it back. Front-card images stay visible alongside the title.

Boxed List Blocks

A boxed list shows a vertical set of items, each in its own coloured container box. There's no numbering or sequence - the boxes read as a set of peers, which makes it the right choice for a list of responsibilities, principles, key points, or "things to remember" where order doesn't matter. It's read-only: learners look at it, they don't interact with it.

Authoring. Click Add Block and choose Boxed List, then:

1. Give the block a title (for example, "Key Responsibilities").
2. Add at least one item. Each item has an optional title (shown in bold at the top of the box) and a body - the body is what the box must contain to be saved. Use Add Item for more; drag the grip handle to reorder.
3. Pick a box colour per item, or use Apply colour to all to set every box at once. The choices are a theme default plus Blue, Teal, Green, Amber, Red, Purple, and Gray.
4. Click Preview to see the boxes laid out as a learner would before saving.

What the learner sees. The title above a stack of coloured boxes, one per item, each showing the item's bold title (if you gave it one) and its body text.

Chart Blocks

A chart block shows a single pie, bar, or line chart over a small set of values. Use it for proportions of a whole (pie), comparisons across categories (bar), or a trend across an ordered axis (line). Charts you import from PowerPoint become editable chart blocks automatically.

Authoring. Click Add Block and choose Chart, then:

1. Give the block a title (required).
2. Choose a chart type - pie, bar, or line.
3. Edit the data as a grid of categories (the labels) and one or more series (the value columns). A pie chart uses the first series only. Add or remove categories and series with the buttons in the grid; leave a value blank for a missing data point.
4. Use Preview to see the chart as a learner would before saving.

What the learner sees. The finished chart, drawn with a fixed colour palette. Keep charts small - around three to eight categories read most clearly; for exact figures the learner must read off precisely, a table on a content slide is better.

Charts inside a content slide. You can also place a chart within a content slide, alongside text and other elements, instead of giving it a slide of its own. In the content editor's toolbar, click the chart button to drop in a chart; click the chart, then the pencil, to open an editor where you choose the chart type and edit the data - the same options as a standalone chart block. This is the right choice when the chart supports a point you're making in the surrounding text; a standalone chart block is better when the chart is the whole slide. Charts imported from PowerPoint that share a slide with other content come in this way automatically.

Clicking the pencil opens the Edit chart editor, where you pick the chart type and edit the categories and series alongside a live preview:

Videos inside a content slide. You can place a video within a content slide, alongside text and other elements, instead of giving it a slide of its own. In the content editor's toolbar, click the video button, then choose where the video comes from:

• YouTube URL - paste a YouTube link. The learner plays it in place; the video stays hosted by YouTube and is subject to its availability and ads.
• Library - pick a video already uploaded to this workspace, or drag in a new one (.mp4, .mov, .webm, up to 500MB). A video you upload is added to the workspace video library and can be reused on any slide in any course in the workspace, so you only upload it once.

Add an optional caption to show a line of text under the player. Click the video, then the pencil, to change the source or the caption later.

This is the right choice when the video supports a point you're making in the surrounding text; a standalone Video or YouTube block is better when the video is the whole slide.

Managing workspace videos. Videos you upload are listed under Settings → Images → Videos, alongside the workspace image library. Each video shows its size, who uploaded it, and how many slides use it. Deleting a video moves it to the workspace trash, where it is kept for 30 days before being permanently deleted - as with images, a video can be deleted even while it is in use, and any slide that used it shows a placeholder in its place.

Resizing blocks in a content slide. Charts, callouts, flow diagrams, code snippets, and videos placed inside a content slide can be resized. Select the block and drag the handles on its edges and corners to make it larger or smaller in either direction; the size you choose is saved with the slide and is what the learner sees. Blocks have a sensible minimum and maximum size, and a block that you size wider than the slide column is automatically capped to fit, so a resized block never overflows the slide. A block you never resize keeps its standard size.

Positioning blocks with text wrap. The same blocks - charts, callouts, flow diagrams, code snippets, and videos - can be floated so the slide text wraps around them. Select the block and use the float buttons that appear in the toolbar: float left (text wraps down the right side), float right (text wraps down the left side), centre (the block is centred on its own line with no wrap), or clear (the block returns to a normal full-width row). A floated block uses the width you've given it by resizing, or a comfortable default width that leaves room for the text to wrap. The position is saved with the slide and is exactly what the learner sees. Inside a two- or three-column slide a floated block wraps the text within its own column, so it never spills into a neighbouring column.

Feedback Blocks

A feedback block collects short reactions from the learner at a point you choose - typically at the end of the course. Each course can have at most one feedback block; the Add Block menu hides the entry once one is already in the course.

Authoring. Click Add Block and choose Feedback, then optionally give the block a title (e.g. "How did this course land?"). The questions themselves are fixed so responses are comparable across every course in the workspace:

• What was the most useful thing you took from this course?
• What did not land for you, or what would you change?
• How likely are you to apply what you learned this week? (1-5)

What the learner sees. The block renders the three questions in a short form; once submitted, the form switches to a read-only confirmation so the learner can't double-submit. Responses are stored against the course version so trainers can review feedback for each release.

Feedback is not assessed

Feedback responses don't contribute to scoring or completion - they're for course improvement, not learner evaluation.

Acknowledgement Blocks

An acknowledgement block asks the learner to read a piece of content and attest that they have read and understood it. It is the right choice for compliance training - ISO, GDPR, HIPAA, internal policies - where you need a signed, dated record that a learner saw the exact wording of a policy.

Authoring. Click Add Block and choose Acknowledgement, then fill in:

• Title (optional) - e.g. "Information Security Policy".
• Text the learner must acknowledge - the full wording of the policy, written in the rich-text editor.
• Require learner to type their full name - leave this on for compliance training; the typed name is stored on the attestation record.
• Button label (optional) - defaults to "I acknowledge".

What the learner sees. The block shows the policy text, a checkbox ("I have read and understood the above"), a name field (if required), and an I acknowledge button that only enables once both inputs are filled. Once submitted, the block switches into a read-only banner showing when they acknowledged it and what name they typed.

What is recorded. Each acknowledgement stores the learner, the slide, the course version, the timestamp, the typed name, and a hash of the exact text the learner saw. Because the hash is frozen at submission, later edits to the block text cannot rewrite history - auditors can verify what wording was actually attested.

Editing an acknowledged block. If you edit the policy text, the course version bumps. Learners who already acknowledged the previous version will be asked to acknowledge the new wording the next time they enter the course.

Voice Narration

Trainers can add AI-generated voice narration to any content block except plain Video and YouTube blocks, which carry their own audio track.

Per-block narration. Click the microphone icon on a block to open the Voice Narration modal. Write your own script or click Generate with AI to produce one from the block's content. The same modal is also reachable from inside the block's own editor - click the Voice button in the editor header (available in every block editor except Video and YouTube) so you don't have to bounce back to the Content tab. The AI script generator reads from the most recently saved block content, so save first if you want the script to reflect your latest edits. When the slide already has audio in more than one language, the Remove button asks whether you want to remove just the language you currently have selected or remove every language and the script. On the Scenario Editor, open the Details tab and click Generate voice - the narration is composed from the scenario's title and description and opens with something like "In this scenario…". If the scenario has no description yet, the button shows an inline notice asking you to add one first. Editing a scenario's title or description clears its audio so the next generation matches the updated text.

Recording or uploading your own voice. If your workspace owner has turned on Recorded voiceovers (Workspace Settings > Features), the Voice Narration modal also shows a Record button and an Upload File button, for any language your workspace has enabled. Record straight from your microphone (up to 10 minutes), or upload a file (mp3, wav, m4a, ogg, or webm, up to 25MB) - either way, a script is optional; you don't need one to record or upload. Whichever you finish most recently is what learners hear immediately. If a slide has both an AI-generated take and a recording for the same language, a small Learners hear: AI / Recorded switch lets you flip between them without losing either one. Click Delete Recording to remove your recording for that language - if an AI-generated version still exists, learners hear that one again.

Bulk narration. Click Generate All Voice in the Content tab toolbar to narrate every content block in the course in one pass - every block type except plain Video and YouTube (which carry their own audio) is included. Each block is narrated from its own content: a boxed list reads its heading and every box's title and body, an accordion or flip card reads each section or card, a chart reads its title and axis labels, and a quiz, matching, or acknowledgement block reads its prompt and content. You can choose Skip existing (only blocks missing audio are processed) or Overwrite (every block is re-generated). Either mode always skips a slide whose learners currently hear a recorded voiceover, so bulk generation never overwrites your own recording. Scenarios that don't have a description yet are reported as failed so you can fill them in and retry. Bulk narration runs in the background - the toolbar shows live progress and you can keep editing the course while it completes; you don't have to keep the tab open.

Per-module narration. To narrate just one module instead of the whole course, click the microphone icon in that module's column header on the Content board. It opens the same modal - with the same Skip existing / Overwrite choice and live progress - but scoped to only the blocks in that module, so you can regenerate a single module after editing it without re-running the rest of the course.

Learner playback. When a block has narration, the course viewer plays it automatically on arrival. Learners can mute from the bottom nav bar, or click the slide content (or press Space while the slide is focused) to pause and resume the voiceover for that slide. Pause resets on every slide change; muting is global for the rest of the session.

AI-generated images

When Aimee generates a course for you (or when you add an image block yourself), slide blocks that still need a picture show an Image needed badge. Click the small image icon on that block to open the Attach image modal, where you can pick from the workspace Library or Upload new. Need an AI-generated picture instead? Generate it from the slide editor (below) and it will appear in the Library on next open.

To generate a fresh image directly from the slide editor, open the slide and click the sparkle icon in the rich-text toolbar - immediately to the right of the Insert image button.

The Generate image with AI modal opens.

• The prompt textarea is seeded with the course description as background context - describe what you want the image to show.
• Pick a style - Photorealistic (default), Illustration, Diagram, or Line art - and submit.
• The preview shows the generated image; click Use this image to insert it at the cursor, or Discard to remove it and try a different prompt. Discarded images are deleted, so they won't litter your image library.
• The image is inserted at the cursor position only - it does not replace any existing slide text. If you have an Image upload required placeholder you want to swap out, delete the placeholder yourself after inserting the image.

Generated images are first-class workspace images: they show up in the Library tab next time you open the image picker, and trainers/admins can manage them on the Course library → Images page.

Learner Brief

The instructions shown to the learner before they launch the simulation. This should explain:

• The situation and context
• Who the learner is playing
• What they need to achieve
• Any constraints or guidelines

Write clear briefs

The learner brief is the only context your trainee gets. Be specific about the course, the learner's role, and success criteria.

Runtime Configuration

A JSON object that configures the AI simulation engine. This may include:

• AI character persona and backstory
• Conversation objectives
• Escalation triggers
• Environmental details

Scenario AI avatar (optional)

On any Scenario block you can turn on an AI avatar that appears alongside the voice runtime - a face that lip-syncs to the AI character as it speaks. Pick a stock avatar from the gallery on the Scenario block's editor, or leave it off to keep the lighter voice-only experience.

The avatar is purely visual; the conversation logic, scoring, and runtime behaviour are unchanged. Learners can mute the audio and the avatar still mouths the words, which can be useful for accessibility scenarios. Avatars add a small amount of usage to the AI Sessions meter while the scenario is active - see Billing for how that's measured.

Course-level AI context

On the course Settings tab, the AI context section lets you describe the course in a way that improves AI-generated content across the editor:

• Intended audience - who the course is for ("new hires in their first 30 days", "regional sales managers"). Aimee tailors generated slides, scoring feedback, and AI image prompts to that audience.
• Out of scope - topics the AI should avoid bringing into the course. Useful when the course deliberately leaves something for a different module.
• Prerequisites - what learners are assumed to know already, so generated content doesn't re-explain it.

These fields are read by AI Course Review, the slide regenerate action, and AI image generation, so the suggestions and reviews come back grounded in your course's audience and scope rather than guessing.

Scoring settings

Setting	Description
Retry Limit	Maximum number of attempts (default: 3)
Pass Threshold	Minimum score to pass (default: 70%)

Grading criteria

Grading criteria live on each scenario - open the scenario from the Content tab and use its Scoring tab. They define how the AI evaluates the roleplay. See Scoring & Rubrics for full details.

Playback

The Playback tab on the editor rail controls how the course plays for learners.

• Enable Autoplay - when on, the course viewer renders an autoplay toggle in the bottom nav bar beside the mute button. Learners opt in per session; the viewer never auto-starts. Once engaged, the course advances automatically:
• Blocks with narration advance when the voice track ends.
• Passive blocks without narration (slides, images, videos, YouTube, resources) advance on a 30-second fallback timer.
• Interactive blocks (quizzes, scenarios, matching, acknowledgements) always wait for the learner to finish.
• Enable Playback Speed (on by default) - exposes a settings cog in the course viewer's bottom bar. Learners can choose between 0.5× and 2× to slow down or speed up narration. Turn this off for compliance or attestation courses where pacing is part of the training record.
• Both settings are version-scoped - changing them and publishing produces a new version only if the overall content has changed, otherwise it overwrites in place.

Previewing a course

Before you publish, click Preview in the Course Editor to see the course exactly as a learner will. The preview opens the learner viewer pinned to your current draft (or the latest published version if there is no draft).

In preview:

• Every slide, image, video, and downloadable resource renders identically to the learner experience.
• Voice narration plays so you can hear pacing, accents, and pronunciation.
• Quizzes show their questions and options, including which answers are correct, so you can sanity-check the answer key.
• Acknowledgements and matching activities show their wording and items but the submit / drag-drop actions are disabled.
• Scenario blocks show their title, description, and grounding image; the AI chat is intentionally not started.

A persistent banner across the top reminds you you're in preview. Click Exit Preview to return to the editor.

Fixing a slide from inside the preview

When you spot something that needs changing, click Edit this slide in the preview banner. It opens that slide in its own editor - no need to exit the preview and hunt for the slide on the content board. The editor then shows a Back to preview control, which brings you back to the slide you were looking at so you can carry on reviewing from the same place.

On anything that isn't a slide - a quiz, a scenario, a matching activity, a resource list - the same button reads Edit this block.

Two kinds of slide can't be edited this way, and the button is greyed out with the reason on hover:

• Branding intro slides, which belong to your workspace's branding intro rather than to this course. Edit them from Workspace Settings → Branding Intro.
• A handful of older slide types that have no editor of their own.

The button doesn't appear at all on courses you've acquired from the marketplace - that content is owned by the publisher and is read-only on your side.

Preview leaves no trace

Preview never starts a real session, never bills against AI usage, and doesn't appear in the workspace activity log. Use it as often as you like.

Accessible to: Trainers and Workspace Owners. Other roles don't see the Preview button, and direct links into the preview URL bounce them back to the catalog.

Publishing a course

Before you can publish, the course must have at least one tag in the Settings tab. Tags are how learners find your course in the catalog and how the workspace organises content - see Workspace Settings → Tags for managing the tag list.

Pre-publish readiness panel

When you click Publish, Aimee runs a quick readiness check on the course and shows the result in a panel before the version is created. The panel groups its findings into two buckets:

• Blockers - things that must be fixed before publication is allowed. Typical blockers are empty slides, missing scenario descriptions, or media blocks without a file attached. Each blocker is a deep-link - click it to jump straight to the offending block.
• Warnings - things that are likely but not certain mistakes (a slide with no voice narration where every other slide has one, an image badge still showing Image needed). Warnings don't block publication; you can publish with a confirmation if you really mean to.

If the course has no tags yet, that's reported as a blocker and clicking it bounces you to the Settings tab so you can add at least one.

Once the readiness check is clean (or you've confirmed past the warnings):

1. Enter a version label (e.g., "v1.0", "v2.0-beta").
2. Confirm the publication.

Published versions are immutable

Once published, a course version cannot be edited. To make changes, create a new version. This ensures learners who were assigned a specific version always see the same content.

Related

• Course Management - drafts, versions, duplicating, templates, Content Library
• Marketplace - publishing your course for other workspaces, or leasing theirs
• Assignments - assigning published courses
• Scoring & Rubrics - rubric configuration
• Trainer role guide - full trainer workflow