Six specialised AI agents work in sequence to transform any document into a complete, pedagogically sound e-learning course — with lessons, quizzes, flashcards, and SCORM export.
Start by providing your content. Learning Catalyst accepts PDF documents, Word files (.docx), plain text, or a public URL. The system supports files up to 50 MB and URLs from most public web pages.
The Extraction Agent reads your document and intelligently splits it into semantic chunks. It identifies headings, key concepts, and logical boundaries — ensuring each lesson covers a coherent topic rather than arbitrary page breaks.
Using your chosen course title and target audience, the Outline Agent generates a full learning architecture: module groupings, lesson titles, and learning objectives aligned to Bloom's Taxonomy. You can customise the number of lessons before generation.
For each lesson, the Content Agent writes structured learning blocks: an introduction, key concept explanations, worked examples, and a summary. Content uses evidence-based instructional design principles including worked examples and dual-coding cues.
The Assessment Agent generates multiple-choice questions for each lesson, with distractors designed to target common misconceptions. A formative quiz appears at the end of every lesson to reinforce retention.
The Flashcard Agent scans each lesson for key terms, definitions, and concepts worth memorising. It produces a spaced-repetition-ready deck that learners can review independently of the main course.
Before delivery, the QA Agent reviews all generated content for consistency, completeness, and alignment with the original source. The course is then packaged and made available for in-platform learning or SCORM export.
Learning Catalyst's prompts are informed by decades of cognitive science and instructional design research.
Content is chunked and sequenced to avoid overloading working memory, following Sweller's CLT principles.
Learning objectives are written at appropriate cognitive levels — from recall to application and analysis.
Segmenting, signalling, and coherence principles guide how content is structured and presented.
Flashcard decks are designed for spaced practice, reinforcing long-term retention beyond the course.
Knowledge checks at the end of each lesson provide immediate feedback and reinforce learning.
Target audience and prior knowledge are factored into tone, complexity, and example selection.
Most courses are generated in 2–5 minutes, depending on document length and the number of lessons requested. Longer documents (50+ pages) may take up to 8 minutes.
PDF, DOCX (Word), TXT, and public URLs. The system extracts text content; images and diagrams within documents are not currently processed.
Yes. Every lesson, quiz question, and flashcard is fully editable in the Course Editor after generation. You can add, remove, or reorder content blocks.
Learning Catalyst supports 65 languages. You select the output language during course setup. English produces the most consistent results; other languages depend on Gemini 2.5 Flash's multilingual capabilities.
Uploaded files are processed in memory and are not stored permanently. Only the generated course content (text) is saved to your account.
SCORM (Sharable Content Object Reference Model) is the standard format for e-learning packages. Export your course as SCORM 1.2 or SCORM 2004 to upload it to any LMS such as Moodle, Canvas, Blackboard, or TalentLMS.
Yes — completely free, forever. There is no subscription, no credit card required, and no monetisation. The platform is a free public service built to democratise e-learning authoring.
Learning Catalyst uses Google Gemini 2.5 Flash for all content generation tasks. The six-agent pipeline orchestrates multiple structured prompts to produce coherent, pedagogically sound output.
Upload a document and have a complete course in under 5 minutes. Free, no credit card required.