.jpg)
We built DreamCycle to address teachers’ overwhelming workload when grading and analyzing student responses. By harnessing AI in a human-centered way, we drastically reduce the time spent on routine tasks while ensuring educators maintain ultimate control. This innovation meets the urgent need for faster, more accurate data cycles that lead to actionable insights and stronger student outcomes.
Teachers administer online assessments via our Illuminate platform, where teachers can upload their students’ complete multiple-choice and constructed-response items. These responses export to EnlightenAI, our AI-scoring tool that uses rubrics, exemplars, and teacher feedback to score submissions accurately and rapidly. Teachers spot-check and confirm the AI’s suggested scores for a small sample, then let it grade the rest — reducing a once 10-hour task to under an hour. The scored data then feeds into a ChatGPT-based system that identifies common mistakes, crafts precise gap statements, and generates full lesson plans for reteaching. At every stage, humans remain in control, validating the AI’s feedback and customizing it to their class context. This approach has allowed DREAM to complete end-to-end data cycles — including analysis and writing reteach lessons—in just days instead of weeks, with strong teacher buy-in and demonstrable improvements in student performance.
We piloted DreamCycle during intensive state test prep and found that teachers instantly saved hours of grading time. Over the past year, weekly use in grades 3–8 has led to improved interim assessment results and deeper data conversations among faculty. Teachers appreciate how quickly they can provide targeted feedback to students. In the next 2–3 years, we plan to introduce DreamCycle to more grade levels and subjects, refine our data analytics for earlier gap detection, and collaborate with partner schools to scale this approach.
Begin by administering assessments or digitizing student work. Export responses in CSV format, then upload to an AI-scoring tool aligned with your rubrics. Validate the AI’s sample scoring and let it handle the rest. Next, feed the results into a data-analysis chatbot to identify gaps and generate reteach plans. For training or implementation guidance, contact DREAM’s Curriculum and Instruction Team.
