AI Student Assessment Trainer

I’m building an AI tool that will automatically evaluate elementary-level student work in Math and Science, then turn those results into clear progress charts teachers can share with parents and use to guide instruction. What I need from you, as an experienced elementary educator, is hands-on guidance and data so the model can learn to judge performance the same way a skilled teacher does. You’ll supply real-world examples of student answers, walk me through the criteria you apply when scoring them, and review the AI’s first drafts of each chart, flagging anything that feels off or misleading. By the end of the engagement the AI should reliably: • score new Math and Science submissions within the same range you would give; • generate a visual progress chart that highlights strengths, pinpoint skills still developing, and suggests the next learning targets. I handle the technical side—prompt engineering, fine-tuning, and deployment. You focus on the pedagogical accuracy so the final model reflects authentic classroom judgment. If that sounds like a good fit, let’s talk timelines and get started.