Turn documents into knowledge graphs where every relationship is a Jev decision backed by evidence.
@richardt830Feed it PDFs, Word files, slides or text. JevGraph reads them locally, finds what they mention, and lines up possible relationships between those things.
Jev then decides which relationship, if any, each pair really has, following a schema you set. Every link in the finished graph points back to the page it came from, so you can check it. Builds on DocJev.
Ask a PostgreSQL database questions in plain English or Chinese, with Jev judging each record and plain SQL doing the joins and totals.
Jev
Python
PostgreSQL
RustSort documents into categories and split bundled PDFs into their parts, using Jev and rules you write in plain English.
JevPythonEnter a payment's details and Jev says whether it looks like fraud, what kind, and whether to approve, challenge or block it.
JevPython
OpenRouterUpload a photo and it sorts the scene into solo, child, couple, family, pet or none, with Jev making the final call.
JevPythonPaste the text of a car related patent in Japanese and Jev picks the single best category for it.
JevPythonOpenRouterUpload a photo, let YOLO draw boxes around the objects you name, and let Jev decide which boxes to keep.
JevPython
