About
This is a multi-agent system designed to generate personalized lesson plans collaboratively with teachers. It operates with a “human-in-the-loop” mechanism, allowing teachers to review and provide feedback which the system uses to reflect and modify the output. The architecture uses a multi-agent system for content curation and sequencing, leveraging LLM reflection to refine outputs based on teacher feedback. It integrates with external knowledge bases for content sourcing and uses templating engines to produce structured, version-controlled lesson plans.
Features
-
Customised Planning: Tailors plans based on subject, learning points, student levels, and duration.
-
Iterative Refinement: Incorporates a feedback loop where teachers can reject or guide the AI for revisions.
-
Structured Output: Delivers finalised lesson plans in a standardised, form-based format.
