Fraya
An AI agent for multimodal e-learning content generation. She uses this documentation as her operational context — consulting it for every decision she makes.
Try asking Fraya
“What are you, and how do you make decisions?”
Agent identity and doc-grounded reasoning
“What's the title and language of course 1550?”
Real-time data access from the course database
“Explain the video generation process.”
Process knowledge retrieved from documentation
“Show me the current Synthesia template-selection prompt.”
Exact prompt contract and stored prompt text
“Evaluate the learning objectives of course 1421.”
Multi-tool reasoning: data + guidelines
The documentation below is what Fraya reads.
Who the learner is, what they need, and the knowledge foundation for the course.
From concept to outline: blueprints, section types, formats, and the full course schema.
Assets
Reusable blocks generated once from the finalized outline — for generation consistency and publishing.
Section Content
Base content layer for all section formats — text, scripts, and structured output.
Native content per language — not translation. System-wide localization rules.
What gets translated, what gets regenerated from scratch.
Guidelines
Standards and rules pulled when needed: course design, personas, language, naming.