Skip to content
FastPII
Developers

Three ways to integrate FastPII into AI applications

Use the open-source SDK for embedded control, the REST API for service access, or the AI Gateway when you want one OpenAI-compatible checkpoint for every model request.

Start FreeView Documentation

Developer-first surface area

Start local, move to APIs, then centralise traffic through the gateway without changing the platform’s core protection model.

Integration paths

SDK, REST API, and Gateway

Choose the integration layer that matches how close you want FastPII to sit to your runtime and provider traffic.

SDK

Embed sensitive data detection and protection directly in Python services with zero external dependencies.

REST API

Call /api/v1/detect, /protect, /validate, and /detectors from any stack that can send JSON.

Gateway

Protect live model traffic through an OpenAI-compatible endpoint with policy and audit controls.

Code examples

Prototype in minutes with Python or cURL

The platform keeps the request contract straightforward so teams can move from local checks to shared services without rewriting everything.

Python
from fastpii import FastPII

client = FastPII()
result = client.detect("PESEL 90010112345 in customer note")
print(result.entities)
cURL
curl https://api.fastpii.com/api/v1/detect \
  -H "Authorization: Bearer $FASTPII_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "IBAN DE89370400440532013000"}'

AI Gateway

Route all model traffic through an OpenAI-compatible proxy that inspects prompts and responses for sensitive data before they reach the provider.

Docs
GitHub
PyPI
Build faster

Pick the integration path that matches your stack today

FastPII lets teams start with the smallest useful surface, then expand to gateway or governance controls as AI traffic grows.

Start FreeView Documentation