Privian
Get Started

Protect prompts before GPT

Replace direct calls to OpenAI with calls to the Privian gateway so prompts are masked before they leave your trust boundary.

Before

ts
// Direct: prompt with PII leaves your trust boundary unmasked.
await fetch("https://api.openai.com/v1/chat/completions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.OPENAI_API_KEY!}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "gpt-4o-mini",
    messages: [{ role: "user", content: "Email john@acme.com about Friday." }],
  }),
});

After

ts
// Via Privian: the masked prompt is forwarded; the response is rehydrated.
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: "Email john@acme.com about Friday.",
    model: "openai/gpt-5.5",
  }),
});

Steps

  1. Create a Privian gateway key in the dashboard.
  2. Add your OpenAI key once as a BYOK provider credential. See BYOK.
  3. Replace the OpenAI endpoint with https://api.privian.io/v1/gateway.
  4. Replace the OpenAI model with a provider-namespaced ID like openai/gpt-5.5.
  5. Read the rehydrated answer from response.

What you gain

  • OpenAI sees masked prompts only.
  • Your application never holds the OpenAI key.
  • Diagnostic counts (entitiesDetected, latency split) without storing the prompt.

Limitations

  • The beta exposes a gateway-native JSON shape (prompt + model), not a drop-in OpenAI SDK replacement.
  • Tool calling, function calling, and structured outputs are not yet exposed via the gateway.

Related