Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

paperless-jev

Docker service that classifies the Paperless-ngx inbox with Jev (or local Clef via Ollama), setting confident fields and queueing the rest.

Projectbeta
View on GitHub ↗

Overview

paperless-jev classifies new documents in a Paperless-ngx inbox with TypeSafe Jev, or locally with Cloudflare's Clef through Ollama: document type, correspondent, storage path, issue date and tags, chosen from your existing Paperless master data. It runs as a Docker container configured entirely through a web UI, triggered by webhook or polling. The README is in German with an English version.

Each answer carries a calibrated confidence: confident values are set and uncertain ones go to a mobile-friendly review queue showing the PDF with Jev's values marked. Modes are dry run, suggestions only, and automatic. A cross-check mode also re-examines existing tags and values and flags or corrects contradictions above a set confidence, and can suggest removing disabled tags. Other features include single-document tests, re-checking by tag, accuracy stats against current filing, LLM-generated titles and descriptions, and phone-based receipt scanning straight into Paperless.

What the decision model decides

TypeQuestionAnswers
ChoiceWhich document type is this?—
ChoiceWhich correspondent sent this document?—
Yes / noDoes this tag apply to the document?(inferred)yes · no

How it works

  1. New document arrives in the Paperless inbox (webhook or polling).
  2. Jev or local Clef picks values from your master data with confidence.
  3. In automatic mode, set confident values and queue the rest for review.
  4. Optionally cross-check existing values and suggest corrections.

Strengths

  • Uses your own master data as options.
  • Three modes from dry run to automatic.
  • Local option via Ollama and Clef.

Caveats

  • Jev mode sends document content to TypeSafe.
  • German-first documentation.

Getting started

Run the Docker container and configure Paperless, Jev or Ollama and mode through the web UI.

At a glance

Type
Project
Status
beta
Runtime
Docker service, web UI
Has code
yes

Use cases and integrations

How curated lists describe it

  • Klassifiziert den Paperless-NGX-Posteingang mit TypeSafe Jev - Docker-Dienst mit Web-UI, Webhook/Polling und Review-Queue.

    hellogumbo/awesome-jev

Frequently asked questions

What does Jev decide?

Document type, correspondent, storage path, date and tags for each new Paperless document.

Can it run without the cloud?

Yes, with Cloudflare's Clef through Ollama.

What happens to uncertain results?

They go to a review queue instead of being applied.

Related projects

  • MultiTool Office Next

    Windows office workbench with search, OCR and batch renaming, where Jev optionally helps pick extracted contract fields.

    ProjectbetaRust
  • SiYuan

    Privacy-first personal knowledge management system, listed as integrating Jev for document classification and tag routing.

    PluginproductionTypeScript
  • tax-doc-classifier

    Classifies tax PDF pages into 261 IRS forms and 7 page kinds with one Jev request per page and a confidence gate.

    ProjectproductionTypeScript
  • doc-router

    Rust tool that asks Jev which PDF pages genuinely need OCR, extracting text pages locally and sending only the rest to OCR.

    ProjectbetaRust
  • docket

    Go terminal app that files documents using five Jev questions: category, sensitivity, urgency, whether action is needed and action state.

    ProjectbetaGo
  • jev-fill-pdf

    Fills Japanese government PDF forms by asking Jev which profile item each unlabeled field is; personal values stay in the browser.

    ServicebetaJavaScript