Privian
Get Started

Add PII masking to OpenAI

Add a PII masking step in front of an existing OpenAI integration without changing how your application reasons about the response.

Migration shape

  1. Store your OpenAI key once as a BYOK credential in Privian. Remove it from your application environment.
  2. Create a Privian gateway key. Set it as PRIVIAN_API_KEY.
  3. Change the endpoint to https://api.privian.io/v1/gateway and the model to openai/<model-id> (e.g. openai/gpt-5.5).
  4. Send { prompt, model } and read response.

Python

python
import os, requests

def ask(prompt: str) -> str:
    r = requests.post(
        "https://api.privian.io/v1/gateway",
        headers={"Authorization": f"Bearer {os.environ['PRIVIAN_API_KEY']}"},
        json={"prompt": prompt, "model": "openai/gpt-5.5"},
        timeout=30,
    )
    r.raise_for_status()
    return r.json()["response"]

TypeScript

ts
export async function ask(prompt: string): Promise<string> {
  const res = await fetch("https://api.privian.io/v1/gateway", {
    method: "POST",
    headers: {
      Authorization: `Bearer ${process.env.PRIVIAN_API_KEY!}`,
      "Content-Type": "application/json",
    },
    body: JSON.stringify({ prompt, model: "openai/gpt-5.5" }),
  });
  if (!res.ok) {
    const { error } = await res.json();
    throw new Error(`privian ${error.code}`);
  }
  const { response } = await res.json();
  return response;
}

Gotchas

  • The gateway accepts a single prompt string today, not the OpenAI messages array. Concatenate roles into a single prompt or keep system instructions out of user-controlled fields.
  • Streaming uses artificial chunking over the rehydrated response; it is not native token streaming.
  • Function calling / tool calling / JSON-mode are not yet exposed via the gateway.

Related