Routing

Regional Lock

Overview

Regional Lock is a powerful feature on slng.ai that allows organizations to strictly control which AI models are used for processing requests, based on the origin or attributes of each call. This ensures that your data is only processed in approved geographic regions, according to your company's compliance, privacy, or operational requirements.

With Regional Lock, you can:

  • Map incoming API calls to specific models or infrastructure in designated regions
  • Define sophisticated rules or policies for routing (e.g., by user, org, country, or request type)
  • Prevent data from being processed in unapproved or risky jurisdictions
  • Minimize compliance exposure and support regulatory requirements (GDPR, HIPAA, etc.)

How It Works

Regional Lock lets you define a set of rules or policies that determine:

  • Which models are eligible to handle a request
  • Where (in which region) those models are allowed to run
  • Fallback behavior if no eligible model is available in the preferred region

This can be as simple as mapping all requests from a certain country to a specific region, or as advanced as using custom logic based on user, org, or request metadata.

Example Policy

JSONCode
{ "rules": [ { "if": { "country": "DE" }, "then": { "region": "EU", "models": ["whisper", "xtts-v2"] } }, { "if": { "org": "acme-corp" }, "then": { "region": "us-west", "models": ["orpheus"] } }, { "if": { "default": true }, "then": { "region": "us-east" } } ] }

Why Use Regional Lock?

  • Compliance: Ensure data never leaves approved regions, supporting GDPR, HIPAA, and other regulations
  • Security: Reduce risk of accidental data residency violations
  • Operational Control: Route requests to regions with optimal latency or cost
  • Custom Logic: Apply rules based on user, org, country, or any request attribute

Example Request

JSONCode
POST /tts Content-Type: application/json { "text": "Lock this to the EU region only.", "language": "en-GB", "region": "EU" }

Best Practices

  • Define clear, explicit rules for all sensitive or regulated data flows
  • Use org-level and request-level overrides for maximum flexibility
  • Regularly review and update your region lock policies as regulations or business needs change
  • Monitor for fallback events to ensure no data is processed outside intended regions

Limitations

  • Some models may not be available in all regions
  • Regional capacity may fluctuate
  • Advanced rule sets may require coordination with your account team


Last updated: June 2025

Last modified on