[Connectors][GenAI] Inference Service Kibana connector

.github/CODEOWNERS

@@ -1639,6 +1639,11 @@ x-pack/test/security_solution_cypress/cypress/tasks/expandable_flyout  @elastic/
 /x-pack/plugins/stack_connectors/server/connector_types/gemini @elastic/security-generative-ai @elastic/obs-ai-assistant @elastic/appex-ai-infra
 /x-pack/plugins/stack_connectors/common/gemini @elastic/security-generative-ai @elastic/obs-ai-assistant @elastic/appex-ai-infra
 
+# Inference API
+/x-pack/plugins/stack_connectors/public/connector_types/inference @elastic/appex-ai-infra
+/x-pack/plugins/stack_connectors/server/connector_types/inference @elastic/appex-ai-infra
+/x-pack/plugins/stack_connectors/common/inference @elastic/appex-ai-infra
+
 ## Defend Workflows owner connectors
 /x-pack/plugins/stack_connectors/public/connector_types/sentinelone @elastic/security-defend-workflows
 /x-pack/plugins/stack_connectors/server/connector_types/sentinelone @elastic/security-defend-workflows

docs/management/action-types.asciidoc

@@ -28,6 +28,10 @@ a| <<gemini-action-type,{gemini}>>
 
 | Send a request to {gemini}.
 
+a| <<inference-action-type,{inference}>>
+
+| Send a request to {inference}.
+
 a| <<email-action-type,Email>>
 
 | Send email from your server.

docs/management/connectors/action-types/inference.asciidoc

@@ -0,0 +1,121 @@
+[[inference-action-type]]
+== {inference} connector and action
+++++
+<titleabbrev>{inference}</titleabbrev>
+++++
+:frontmatter-description: Add a connector that can send requests to {inference}.
+:frontmatter-tags-products: [kibana] 
+:frontmatter-tags-content-type: [how-to] 
+:frontmatter-tags-user-goals: [configure]
+
+
+The {inference} connector uses {es} client to send requests to {inference}. The connector uses the <<execute-connector-api,run connector API>> to send the request.
+
+[float]
+[[define-inference-ui]]
+=== Create connectors in {kib}
+
+You can create connectors in *{stack-manage-app} > {connectors-ui}*.  For example:
+
+[role="screenshot"]
+image::management/connectors/images/inference-connector.png[{inference} connector]
+// NOTE: This is an autogenerated screenshot. Do not edit it directly.
+
+[float]
+[[inference-connector-configuration]]
+==== Connector configuration
+
+{inference} connectors have the following configuration properties:
+
+Name::      The name of the connector.
+Service::   The {inference} supported provider service.
+Task type:: The {inference} task type, it depends to the selected service.
+Inference ID:: The {inference} ID.
+Provider configuration:: Settings for service configuration.
+Provider secrets:: Configuration for authentication.
+Task type configuration:: Settings for task type configuration.
+
+[float]
+[[inference-action-configuration]]
+=== Test connectors
+
+You can test connectors with the <<execute-connector-api,run connector API>> or
+as you're creating or editing the connector in {kib}. For example:
+
+[role="screenshot"]
+image::management/connectors/images/inference-completion-params.png[{inference} params test]
+// NOTE: This is an autogenerated screenshot. Do not edit it directly.
+
+The {inference} actions have the following configuration properties. Properties depends to the selected task type.
+
+[float]
+[[inference-connector-perform-completion]]
+=== Connector perform completion task
+
+The following example performs a completion on the example question.
+Input::      The text on which you want to perform the inference task. For example:
++
+[source,text]
+--
+
+{
+  input: 'What is Elastic?'
+}
+--
+
+[float]
+[[inference-connector-perform-text-embedding]]
+=== Connector perform text embedding
+
+The following example performs text embedding.
+Input::      The text on which you want to perform the inference task. For example:
++
+[source,text]
+--
+
+{
+  input: 'The sky above the port was the color of television tuned to a dead channel.',
+  task_settings: {
+    input_type: 'ingest
+  }
+}
+--
+Input type::      An optional string that overwrites the connector's default model.
+
+[float]
+[[inference-connector-perform-rerank]]
+=== Connector perform rerank task
+
+The following example performs reranking on the example input.
+Input::      The text on which you want to perform the inference task. Should be a string array. For example:
++
+[source,text]
+--
+
+{
+  input: ['luke', 'like', 'leia', 'chewy', 'r2d2', 'star', 'wars'],
+  query: 'star wars main character'
+}
+--
+Query::      The search query text.
+
+[float]
+[[inference-connector-perform-sparse-embedding]]
+=== Connector perform sparse embedding task
+
+The following example performs sparse embedding on the example sentence.
+Input::      The text on which you want to perform the inference task. For example:
++
+[source,text]
+--
+
+{
+  input: 'The sky above the port was the color of television tuned to a dead channel.'
+}
+--
+
+[float]
+[[inference-connector-networking-configuration]]
+=== Connector networking configuration
+
+Use the <<action-settings, Action configuration settings>> to customize connector networking configurations, such as proxies, certificates, or TLS settings. You can set configurations that apply to all your connectors or use `xpack.actions.customHostSettings` to set per-host configurations.

docs/management/connectors/index.asciidoc

@@ -4,6 +4,7 @@ include::action-types/crowdstrike.asciidoc[leveloffset=+1]
 include::action-types/d3security.asciidoc[leveloffset=+1]
 include::action-types/email.asciidoc[leveloffset=+1]
 include::action-types/gemini.asciidoc[leveloffset=+1]
+include::action-types/inference.asciidoc[leveloffset=+1]
 include::action-types/resilient.asciidoc[leveloffset=+1]
 include::action-types/index.asciidoc[leveloffset=+1]
 include::action-types/jira.asciidoc[leveloffset=+1]

docs/settings/alert-action-settings.asciidoc

@@ -283,6 +283,7 @@ A configuration URL that varies by connector:
 --
 * For an <<bedrock-action-type,{bedrock} connector>>, specifies the {bedrock} request URL.
 * For an <<gemini-action-type,{gemini} connector>>, specifies the {gemini} request URL.
+* For an <<inference-action-type,{inference} connector>>, specifies the Elastic {inference} request.
 * For a <<openai-action-type,OpenAI connector>>, specifies the OpenAI request URL.
 * For a <<resilient-action-type,{ibm-r} connector>>, specifies the {ibm-r} instance URL.
 * For a <<jira-action-type,Jira connector>>, specifies the Jira instance URL.

oas_docs/examples/get_connector_types_generativeai_response.yaml

@@ -31,3 +31,14 @@ value:
     supported_feature_ids:
       - generativeAIForSecurity
     is_system_action_type: false
+  - id: .inference
+    name: Inference API
+    enabled: true
+    enabled_in_config: true
+    enabled_in_license: true
+    minimum_license_required: enterprise
+    supported_feature_ids:
+      - generativeAIForSecurity
+      - generativeAIForObservability
+      - generativeAIForSearchPlayground
+    is_system_action_type: false

x-pack/packages/kbn-elastic-assistant/impl/mock/connectors.ts

@@ -39,6 +39,16 @@ export const mockActionTypes = [
     isSystemActionType: true,
     supportedFeatureIds: ['generativeAI'],
   } as ActionType,
+  {
+    id: '.inference',
+    name: 'Inference API',
+    enabled: true,
+    enabledInConfig: true,
+    enabledInLicense: true,
+    minimumLicenseRequired: 'enterprise',
+    isSystemActionType: true,
+    supportedFeatureIds: ['generativeAI'],
+  } as ActionType,
 ];
 
 export const mockConnectors: AIConnector[] = [

x-pack/plugins/actions/common/connector_feature_config.ts

@@ -46,7 +46,7 @@ const compatibilityGenerativeAIForObservability = i18n.translate(
 const compatibilityGenerativeAIForSearchPlayground = i18n.translate(
   'xpack.actions.availableConnectorFeatures.compatibility.generativeAIForSearchPlayground',
   {
-    defaultMessage: 'Generative AI for Search Playground',
+    defaultMessage: 'Generative AI for Search',
   }
 );
 

x-pack/plugins/actions/docs/openapi/components/schemas/connector_types.yaml

@@ -4,6 +4,7 @@ description: The type of connector. For example, `.email`, `.index`, `.jira`, `.
 enum:
   - .bedrock
   - .gemini
+  - .inference
   - .cases-webhook
   - .d3security
   - .email

x-pack/plugins/actions/docs/openapi/components/schemas/inference_config.yaml

@@ -0,0 +1,22 @@
+title: Connector request properties for an Inference API connector
+description: Defines properties for connectors when type is `.inference`.
+type: object
+required:
+  - provider
+  - taskType
+properties:
+  provider:
+    type: string
+    description: The Inference API service provider.
+  taskType:
+    type: string
+    description: The Inference task type supported by provider.
+  providerConfig:
+    type: object
+    description: The provider settings.
+  taskTypeConfig:
+    type: object
+    description: The task type settings.
+  inferenceId:
+    type: string
+    description: The task type settings.

x-pack/plugins/actions/docs/openapi/components/schemas/inference_secrets.yaml

@@ -0,0 +1,9 @@
+title: Connector secrets properties for an AI Connector
+description: Defines secrets for connectors when type is `.inference`.
+type: object
+required:
+  - providerSecrets
+properties:
+  providerSecrets:
+    type: object
+    description: The service account credentials. The service account could have different type of properties to encode.