Intrace Intelligence API

Deployment

Configuration

Runtime configuration notes for customer deployments

The API is deployed as a containerized service. Configure accepted API keys, optional Redis/Valkey-backed shared state, connector runtime settings, social-match evaluator settings, and required PII expansion AI settings through environment variables.

Common variables:

  • INTELLIGENCE_API_API_KEYS or INTELLIGENCE_API_KEYS_JSON: API keys accepted by the service.
  • INTELLIGENCE_API_REDIS_URL: optional Redis/Valkey URL for shared job lookup and rate limiting.
  • INTELLIGENCE_API_JOB_STORE_BACKEND: memory or redis.
  • INTELLIGENCE_API_RATE_LIMIT_BACKEND: memory or redis.
  • INTELLIGENCE_API_SOCIAL_MATCH_EVALUATOR_BASE_URL: optional external evaluator base URL.
  • INTELLIGENCE_API_SOCIAL_MATCH_EVALUATOR_API_KEY: optional external evaluator API key.
  • INTELLIGENCE_API_SOCIAL_MATCH_EVALUATOR_MODEL: optional evaluator model name.
  • INTELLIGENCE_API_PII_EXPANSION_AI_API_KEY or OPENROUTER_API_KEY: required for PII expansion semantic dedupe and validation.
  • INTELLIGENCE_API_PII_EXPANSION_AI_MODEL: PII expansion AI model name.
  • INTELLIGENCE_API_PII_EXPANSION_AI_TIMEOUT_SECONDS: PII expansion AI timeout.

For multi-replica deployments, use Redis/Valkey-backed job storage and rate limiting so job lookup and request budgets are shared across instances.