ExamAI is an offline AI study assistant that lets students ask questions directly from their PDF notes. It uses local embeddings, ChromaDB, and Ollama to keep study data on the user's computer. It supports multiple PDFs, document switching, exam answers for 2M, 5M, 8M and 10M questions, and full-document summaries.
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Maker
š
Hey Product Hunt š
I built ExamAI because I wanted a simpler way to study from my own notes without sending them to a cloud AI service.
ExamAI lets you upload your PDF study material and ask questions directly from it. It uses local AI, embeddings, and vector search to keep the workflow offline.
A few things I built into it:
⢠Multiple PDF support with document switching
⢠Exam Mode for 2M, 5M, 8M and 10M answers
⢠Full-document summary mode
⢠Local ChromaDB vector storage
⢠Local Ollama inference
⢠No external AI API required
The biggest challenge was making the retrieval system work reliably across multiple documents while keeping each document's context isolated.
I'd really appreciate feedback from the Product Hunt community ā especially on the product positioning, UX, and what feature you think I should build next.
Thanks for checking out ExamAI! š