The AI Gateway
In Action

How the Request Pipeline Works

When an employee sends a prompt to any AI tool, the request first hits the Tenlines gateway. The gateway inspects the full prompt body for personally identifiable information (PII) such as names, email addresses, phone numbers, social security numbers, and credit card numbers. It also detects API keys, authentication tokens, and other credentials that could cause a security incident if leaked.

Detected sensitive data is handled according to your organization's policy. You can redact it entirely — replacing values with safe placeholders like [EMAIL_REDACTED]. You can tokenize it — swapping real values for reversible tokens so AI responses can be de-tokenized back to the original. Or you can encrypt it with AES-256-GCM, where only your organization holds the decryption key.

Beyond PII, Tenlines also checks prompts against your registered proprietary code. Using semantic embedding similarity, it detects when employees paste source code, internal documentation, or trade secrets that should never reach an external AI provider.

After scrubbing, the clean request is forwarded to the AI provider. The response streams back through the gateway, where it can be de-tokenized or decrypted before reaching the employee. The entire interaction — including what was detected, what action was taken, and which user initiated it — is written to an audit log.