Many teams self-host because they cannot send sensitive data to a hosted model like GPT or Claude. Protecto changes that. Your real values stay in your environment. The model receives masked values only.
| Before the Gateway | With the Gateway | |
|---|---|---|
| To keep sensitive data private, you had to | self-host, and accept whichever models you could run | Use any model. |
| Model choice was | a security decision | a quality decision |
| Frontier models were | off limits for sensitive work | available, because the model never sees real values |
When you send a prompt to a hosted model, the exposure is not only the prompt you just sent. The session persists on the provider’s side: every earlier message, the context built across turns, the detail a user typed three questions ago. You control none of it. Not retention, not access, not training use, not third-party sharing.
Sensitive information also accumulates differently across a conversation than it appears in one prompt. Nothing in turn four looks sensitive on its own. Turns one through four together identify a patient.
Run a self-hosted chat interface behind the Gateway and the session stays on your infrastructure. The provider only ever holds masked values, on every turn, for the life of the conversation.
Basic masking: The tool gets a surrogate, so exact-match lookups fail and the model answers from nothing.
Protecto: Resolves supported tool arguments at the tool boundary, so approved tool calls get the real values they need.
Basic masking: Redaction markers change token usage, and rescanning at every step adds processing and latency.
Protecto: Uses consistent, format-preserving surrogates. The impact depends on value length, tokenizer, and payload size.
Basic masking: Scan-then-forward needs the full payload before the first model token can go out.
Protecto: Streams the model’s response back after masking.
How to mask sensitive values for the model and still give an approved tool the real value it needs, with a worked tool-call example and a checklist for production.
The detection and masking engine behind the Gateway: how Protecto Vault finds, masks, and controls access to sensitive data across AI applications and enterprise systems.
How a health analytics company built a data product on sensitive medical text while keeping patient data private.
A walk through one LibreChat conversation behind the Gateway: what the user sees, what the model receives, and how tool calls still resolve.
Gateway latency depends on prompt size, entity density, selected model, synchronous versus asynchronous masking, and whether tool arguments require resolution. During deployment, Protecto measures median and tail latency against your representative prompt sizes and workflows.
A local model is a legitimate choice and plenty of good teams have made it. It is also a choice most teams made because it was the only one available, and it costs capability on every task where a frontier model is better. The Gateway does not argue against your local model. What changes is that using a hosted model like GPT, Claude, or Gemini alongside it stops being a security decision.
Yes, because the masking is context-preserving and format-preserving rather than blanket redaction. The model reasons over structurally intact, consistent values instead of blacked-out text. In a deployment at a large bank in the Middle East, semantic accuracy held up well after masking.
Protecto is built primarily for on-prem, VPC, or air-gapped deployment, so the mapping and the masked data stay inside your environment by default. Protecto maintains the authoritative token-to-value mapping in the configured Privacy Vault. The Gateway holds mappings transiently while processing a request, and your chat interface stores encoded replay data used to reconstruct masked history. Unmasking uses a per-user Protecto token. Access to the vault, chat interface artifacts, and operational logs must be restricted.
The detection and masking engine is Vault, and it is already in production at named enterprise customers. What is new is where it runs and what it does on the way back. The Gateway sits in the chat runtime path and resolves a tagged value at the tool boundary whenever it is present and the per-user Protecto token permits it, so tool calls resolve and the user sees real data while the model never held a real value.
Protecto Gateway is designed to detect and mask supported entity-based sensitive data, including common personal, health, payment, account, government, and customer identifiers, before protected chat content reaches the model provider.
The Gateway is decoupled from the chat interface, so nothing in the architecture is tied to one UI. LibreChat is the integration available today. Open WebUI and other interfaces are available on request, and the fastest path is a short call where we look at your instance.
No. Attaching raw file bytes directly to the model provider is not supported. Files are read through the chat interface and masked before their content enters the prompt, which is what keeps the control in place.
No. The Gateway does not itself authenticate the chat session. It trusts caller-supplied identity and relies on network placement, the same as the systems in front of it. It does work with AD and Okta: you keep each user’s masking and unmasking access in AD or Okta groups, and the Gateway issues a Protecto auth key with that access at login.
Protecto detects 50+ built-in entity types across personal, health, and payment data, plus custom entities you define. Detection works in English, Arabic, French, German, Spanish, Dutch, and Japanese, with more languages on request.
As SaaS, on-prem, in a private cloud, or air-gapped. Teams that self-host their chat interface usually run Protecto in the same environment, so real values stay inside it.
Protecto holds SOC 2 Type II and ISO 27001. Protecto is HIPAA compliant. When Protecto hosts the service or can access your instance, Protecto signs a BAA.
Book a demo and bring the model and chat interface you use. We run the Gateway on your representative prompts and workflows, including tool calls, and measure latency against them.
A practical guide to privacy controls for AI assistants, RAG, and agents. One workflow traced from request to CRM update, with what each destination may see.
By submitting you agree to Protecto’s privacy policy.
Features that safeguard your data across agents, AI applications, RAG pipelines, MCPs, data lakes, and enterprise systems.
By submitting you agree to Protecto’s privacy policy.