Skip to main content

Regional Streaming Controls

Latency, compliance, and model availability often depend on which region you call. NeuroLink threads the region parameter through the generate/stream stack so you can target specific data centres when working with providers that expose regional endpoints.

Quick Start​

import { NeuroLink } from "@juspay/neurolink";

const neurolink = new NeuroLink();

const result = await neurolink.stream({
  input: { text: "Summarise EMEA incident reports" },
  provider: "bedrock",
  model: "anthropic.claude-sonnet-4-6",
  region: "eu-central-1",
});

for await (const chunk of result.stream) {
  process.stdout.write(chunk);
}

Supported Providers​

ProviderHow to Set RegionDefaults
Amazon BedrockAWS_REGION env, config init, or request region optionus-east-1
Amazon SageMakerSAGEMAKER_DEFAULT_ENDPOINT + AWS_REGION or request regionus-east-1
Google Vertex AIGOOGLE_VERTEX_LOCATION / config init / request regionus-east5
Azure OpenAIDeployment-specific endpoint; use AZURE_OPENAI_ENDPOINT (region encoded)—
LiteLLM pass-throughUse LiteLLM server configuration—

Providers without native region controls ignore the option safely.

CLI Usage​

The CLI reads region information from configuration profiles or provider environment variables.

# Bedrock: ensure AWS credentials + region set
export AWS_REGION=ap-south-1
npx @juspay/neurolink generate "Translate catalog" --provider bedrock

# Vertex AI: switch to Tokyo region for lower latency
export GOOGLE_VERTEX_LOCATION=asia-northeast1
npx @juspay/neurolink stream "Localise onboarding" --provider vertex --model gemini-2.5-pro

# One-off override via shell env
AWS_REGION=eu-west-1 npx @juspay/neurolink stream "Summarise EMEA incidents" --provider bedrock

Run neurolink config init to persist region defaults per provider.

SDK Usage​

const neurolink = new NeuroLink({ enableOrchestration: true });

const result = await neurolink.generate({
  input: { text: "Compile regional latency metrics" },
  provider: "vertex",
  model: "gemini-2.5-pro",
  region: "europe-west4",
  enableEvaluation: true,
});

console.log(result.content, result.provider);

Streaming obeys the same option:

const stream = await neurolink.stream({
  input: { text: "Narrate service availability" },
  provider: "bedrock",
  model: "anthropic.claude-sonnet-4-6",
  region: "eu-central-1",
});

Operational Tips​

Compliance & Data Residency

Use regional routing to comply with data sovereignty requirements (GDPR, HIPAA, etc.). Pin the region parameter to ensure AI processing stays within approved geographical boundaries for sensitive workloads.

Latency Optimization

Co-locate your NeuroLink deployment with your application servers. For example, if your API runs in eu-west-1, set region: "eu-west-1" for Bedrock/Vertex calls to minimize cross-region latency penalties.

  • Compliance – ensure the requested region is enabled for the model (e.g., Anthropic via Vertex only supports us regions).
  • Latency – co-locate with your application servers to avoid cross-region penalties.
  • Fallbacks – when orchestration re-routes to a provider that ignores region, the call completes but logs a warning.
  • Credentials – AWS requests still require valid IAM credentials; Vertex needs service account rights in the target location.

Troubleshooting​

SymptomFix
Invalid region formatUse standard IDs (us-east-1, asia-northeast1).
Model not available in regionSwitch to a supported region or change model (see provider console).
Credential error after region changeRe-run neurolink config init so stored credentials match the new region.
High latency on fallback providerDisable orchestration or pin a provider/model explicitly.