Upload a rental agreement and see risky clauses explained in plain language
Intermediate14h1–3
PDF.jsLLMNext.js
16/20Document Q&A
Lets people ask questions about long documents instantly.
Intermediate24h1–3 people16/20 fit
Build Document Q&A for students, researchers, and professionals in ~24h with 1–3 people. Stack: Next.js + RAG + LLM.
Why it wins: Classic strong demo that judges recognize.
Build flow
01 · Problem
Long PDFs are hard to search when you need a specific answer.
02 · Solution
Upload a PDF. Ask questions and get answers with citations.
03 · Build
- 1Extract + chunk the PDF
- 2Embed chunks
- 3Retrieve + answer
- 4Show citations
04 · Demo
Upload a short PDF → ask one question → show answer + source.
MVP checklist
- PDF upload
- Chat-style Q&A
- Cited sources
- Highlight in document
Tech stack
| Tech | Role | |
|---|---|---|
| Next.js | Frontend + API routes | Web |
| RAG | Retrieval over your docs | AI |
| LLM | Language model | AI |
Reference repos
Open-source projects you can build on. Links checked 2026-10-31.
Related ideas
More AIAnswer customer questions on price, stock and delivery from your own catalog
Intermediate16h1–3RAGVector DBLLM16/20Extract tables from messy PDFs into clean, checked CSV files
Intermediate14h1–2MarkItDownPythonLLM16/20