It differs from using an AI chatbot for writing, in that it has been designed with specific constructs in mind. Particularly to offer a structured evaluation using the CEFR Framework as guard rails to optimize learning, rather than a free flowing conversation with an LLM using ambiguous parameters. Evaluation is visualized as a matrix rubric which students can immediately comprehend and understand the linguistic areas that they are strongest/weakest in.

Description: The primary purpose of this tool is to offer students a more intensive way to exercise their writing skills without increasing teacher burden. Its database is managed via Claude and updated regularly with added sample essays. The tool does not use LLM models to evaluate essays, instead relying on algorithmic heuristics to measure essay length, lexical and sentence complexity as well as coherence and flow (transition words, complete thoughts/sentences, hedging) and also penalizes redundancy.