Dewey – Ingest docs, search semantically, get cited AI answers
Show HN: Dewey – Ingest docs, search semantically, get cited AI answers
Based, Dewey – Ingest docs, search semantically, get cited AI answers works in show hn launch.
Profile
Flat chunking throws away document structure. A PDF isn’t a bag of paragraphs. It has sections, subsections, and a hierarchy that carries meaning. An agent that can’t navigate that structure can’t do serious research.I ran into this building RAG over scientific literature. The standard approach (embed chunks, find top-k, generate) works fine for simple Q&A but falls apart when you need real research depth: multi-hop reasoning across papers, synthesizing conflicting results, tracing a finding back to the exact passage in a methods section. The problem wasn’t the models.Dewey treats documents, sections, and chunks as first-class API primitives. The section manifest (full heading hierarchy with
The company sits in the show hn launch segment, where it competes on its core product or service.
The company is listed at the Show HN launch stage of development.
| Sector | Show HN launch |
|---|---|
| Stage | Show HN launch |
Website: meetdewey.com
Browse related
Create a free account