Nexus · Ingest pipeline

Add knowledge artifacts

Nexus pulls text from PDFs, images with OCR when needed, and plain text uploads. Materials are chunked with overlap, embedded, and persisted for semantic retrieval later.

Upload·Extract / OCR·Chunk 500 / 50·Embed 768-D·pgvector

Drag a supported file here or browse from disk.

Select a filePDF, images, Markdown, TXT · up to ~20 MB depending on tier

Add knowledge artifacts

File drop zone