Documentation

Everything you need to classify files with Veriafy while maintaining complete privacy.

Quick Start Guide

Start classifying files with Veriafy in minutes

1. Install Veriafy

Install via our installation script or package manager:

curl -fsSL https://get.veriafy.com/install.sh | sh

2. Initialize Veriafy

Set up your local environment:

veriafy init

3. Download a Model

Pull a classification model from the marketplace:

veriafy pull fraud-detection

4. Classify a File

Run classification on any file — Veriafy never sees the content:

veriafy classify document.pdf --model fraud-detection

Output: vector_id, categories, confidence scores, and recommended action

5. Use the Python SDK

Integrate Veriafy into your applications:

from veriafy import Veriafy

client = Veriafy()  # Local mode, no API key needed

result = client.classify("document.pdf", model="fraud-detection")

print(result.vector_id)    # Unique hash identifier
print(result.categories)   # {'fraud': 0.92, 'legitimate': 0.08}
print(result.action)       # 'flag'

GitHub

Source code, examples, and contributions

Discord

Join our developer community

API Reference

Complete REST API documentation