EdTech
AI-Use Transparency Log
Students record how they used AI on an assignment for honest disclosure
Intermediate12h1–2
Next.jsLLMPostgres
17/20Rubric-Based Essay Feedback
Instant feedback on drafts mapped to the teacher's rubric.
Intermediate14h1–2 people17/20 fit
Build Rubric-Based Essay Feedback for teachers and students in ~14h with 1–2 people. Stack: LLM + Next.js + Postgres.
Why it wins: Supports teachers instead of replacing them.
Build flow
01 · Problem
Students wait days for feedback and teachers cannot comment on every draft.
02 · Solution
Teachers upload a rubric, students submit drafts and an LLM gives criterion-level comments with quoted evidence, never a final grade.
03 · Build
- 1Model rubrics
- 2Prompt per criterion
- 3Show comments
- 4Track revisions
04 · Demo
Submit a draft and show comments per rubric row.
MVP checklist
- Rubric upload
- Criterion feedback
- Quoted evidence
- Revision history
Tech stack
| Tech | Role | |
|---|---|---|
| LLM | Language model | AI |
| Next.js | Frontend + API routes | Web |
| Postgres | Relational database | Data |
Related ideas
More EdTechStudents review each other's code anonymously with a guided checklist
Intermediate14h2–3Next.jsPostgresLLM16/20Generate a lesson plan aligned to a curriculum standard with activities and checks
Intermediate12h1–2LLMNext.js17/20