Skip to content

API Gateway

Put your own APIs behind the same gateway.

Airelay isn't only for AI traffic. Publish your own services at https://ai.noviqent.co.uk/gw/<your-org>/… with authentication, rate limiting, transformation, health-checked load balancing and logging — configured from the dashboard, no gateway to run yourself.

Services, routes, done

Describe the backend you already have as a service, attach routes under your organisation's path, and Airelay starts proxying — with every plugin you attach applied on the way through.

terminal
# Your service "orders", route /v1/orders, key-auth enabled
curl https://ai.noviqent.co.uk/gw/acme/v1/orders \
  -H "apikey: $PARTNER_KEY"

# → authenticated, rate-limited, logged,
#   load-balanced across your healthy upstream targets
  • Services point at a URL or at one of your own load-balanced upstreams.
  • Routes match on path (prefix or regex), method and headers under /gw/<your-org>/.
  • Plugins attach to a service, a route, a consumer or a consumer group.
  • Every organisation is isolated: your routes, consumers and plugins can't see or be reached by another tenant's.
  • Outbound targets are checked so a service can't be pointed at private or internal network addresses.

Consumers, credentials and groups

Model who calls your APIs. Give each consumer the credential types you need, group them, and apply plugins — like a higher rate limit for a partner tier — to a whole group at once.

  • API keys
  • JWT
  • OAuth 2.0
  • Basic auth
  • HMAC
  • LDAP
  • ACL groups

Consumer groups

Add consumers to a group and attach plugins to the group — membership changes apply automatically, and ACL rules can allow or deny a whole group.

Per-consumer policy

Different limits, transformations or logging for a specific consumer, without touching the route everyone else uses.

Keys you control

Issue, rotate and revoke credentials from the dashboard; secrets can reference values held in your Noviqent Vault space instead of being pasted in.

48 plugins, configured from forms

Each plugin's settings are rendered from its own schema, so what you can configure in the dashboard is exactly what the gateway supports. Highlights:

Authentication

  • Key auth
  • Basic auth
  • JWT
  • OAuth 2.0
  • HMAC
  • LDAP
  • ACL

Security

  • CORS
  • IP restriction
  • Bot detection
  • Request size limiting

Traffic control

  • Rate limiting
  • Response rate limiting
  • Proxy caching
  • Request termination

Transformation

  • Request transformer
  • Response transformer
  • Correlation ID
  • gRPC-Web

Logging & tracing

  • HTTP
  • TCP
  • UDP
  • Datadog
  • OpenTelemetry
  • Zipkin
  • StatsD

Serverless

  • AWS Lambda
  • Azure Functions

The AI plugins — PII redaction, guardrails, semantic cache and more — can be attached to your own API routes too.

Upstreams, load balancing and health checks

Point a service at an upstream and spread traffic across several targets.

Load balancing

Weighted round-robin, least-connections or consistent hashing across the targets you add to an upstream.

Active health checks

Probe each target on a schedule and take it out of rotation when it stops answering — then bring it back when it recovers.

Passive health checks

Watch real traffic for errors and timeouts and stop routing to a target that's failing, before the next probe notices. See target health live in the dashboard.

Certificates, keys and secrets

Manage the cryptographic material your APIs need alongside the routes that use it.

  • TLS certificates and SNIs, and CA certificates for verifying clients or upstreams.
  • Keys and key sets (JWK / PEM) for JWT signing and verification.
  • Vault-backed secrets: reference a value stored in your Noviqent Vault space from any plugin setting — the dashboard shows the reference, never the value.

A developer portal for your APIs

Upload an OpenAPI spec per service and publish a public portal page for your organisation — so the people calling your APIs can read the docs without an account.

Your AI traffic and your APIs, one gateway

Everything here lives in the same dashboard as your AI gateway — same team roles, same audit log, same analytics.