diff --git a/.env.sample b/.env.sample
index a9ed338d..f1dfa190 100644
--- a/.env.sample
+++ b/.env.sample
@@ -14,8 +14,10 @@ LEON_LLM=true
 LEON_LLM_PROVIDER=local
 # LLM provider API key (if not local)
 LEON_LLM_PROVIDER_API_KEY=
-# Enable/disable LLM Natural Language Generation
+# Enable/disable LLM natural language generation
 LEON_LLM_NLG=true
+# Enable/disable LLM Action Recognition
+LEON_LLM_ACTION_RECOGNITION=true
 
 # Time zone (current one by default)
 LEON_TIME_ZONE=
diff --git a/server/src/core/http-server/api/llm-inference/post.ts b/server/src/core/http-server/api/llm-inference/post.ts
index 3b986027..e6a8572d 100644
--- a/server/src/core/http-server/api/llm-inference/post.ts
+++ b/server/src/core/http-server/api/llm-inference/post.ts
@@ -7,6 +7,7 @@ import { SummarizationLLMDuty } from '@/core/llm-manager/llm-duties/summarizatio
 import { TranslationLLMDuty } from '@/core/llm-manager/llm-duties/translation-llm-duty'
 import { ParaphraseLLMDuty } from '@/core/llm-manager/llm-duties/paraphrase-llm-duty'
 import { ChitChatLLMDuty } from '@/core/llm-manager/llm-duties/chit-chat-llm-duty'
+import { ActionRecognitionLLMDuty } from '@/core/llm-manager/llm-duties/action-recognition-llm-duty'
 import { LLM_MANAGER } from '@/core'
 
 interface PostLLMInferenceSchema {
@@ -19,6 +20,7 @@ interface PostLLMInferenceSchema {
 }
 
 const LLM_DUTIES_MAP = {
+  [LLMDuties.ActionRecognition]: ActionRecognitionLLMDuty,
   [LLMDuties.CustomNER]: CustomNERLLMDuty,
   [LLMDuties.Summarization]: SummarizationLLMDuty,
   [LLMDuties.Translation]: TranslationLLMDuty,
diff --git a/server/src/core/llm-manager/llm-duties/action-recognition-llm-duty.ts b/server/src/core/llm-manager/llm-duties/action-recognition-llm-duty.ts
new file mode 100644
index 00000000..d2060889
--- /dev/null
+++ b/server/src/core/llm-manager/llm-duties/action-recognition-llm-duty.ts
@@ -0,0 +1,81 @@
+import {
+  type LLMDutyParams,
+  type LLMDutyResult,
+  LLMDuty
+} from '@/core/llm-manager/llm-duty'
+import { LogHelper } from '@/helpers/log-helper'
+import { LLM_MANAGER, LLM_PROVIDER } from '@/core'
+import { LLM_THREADS } from '@/core/llm-manager/llm-manager'
+import { LLMProviders } from '@/core/llm-manager/types'
+import { LLM_PROVIDER as LLM_PROVIDER_NAME } from '@/constants'
+
+interface ActionRecognitionLLMDutyParams extends LLMDutyParams {}
+
+const JSON_KEY_RESPONSE = 'action_name'
+
+export class ActionRecognitionLLMDuty extends LLMDuty {
+  protected readonly systemPrompt = `You are an AI expert in intent classification and matching.
+You look up every utterance sample and description. Then you return the most probable intent (action) to be triggered based on a given utterance.
+If the intent is not listed, do not make it up yourself. Instead you must return { "${JSON_KEY_RESPONSE}": "not_found" }.`
+  protected readonly name = 'Action Recognition LLM Duty'
+  protected input: LLMDutyParams['input'] = null
+
+  constructor(params: ActionRecognitionLLMDutyParams) {
+    super()
+
+    LogHelper.title(this.name)
+    LogHelper.success('New instance')
+
+    this.input = params.input
+  }
+
+  public async execute(): Promise<LLMDutyResult | null> {
+    LogHelper.title(this.name)
+    LogHelper.info('Executing...')
+
+    try {
+      const prompt = `Utterance: "${this.input}"`
+      const completionParams = {
+        systemPrompt: this.systemPrompt,
+        data: {
+          [JSON_KEY_RESPONSE]: {
+            type: 'string'
+          }
+        }
+      }
+      let completionResult
+
+      if (LLM_PROVIDER_NAME === LLMProviders.Local) {
+        const { LlamaChatSession } = await Function(
+          'return import("node-llama-cpp")'
+        )()
+
+        const context = await LLM_MANAGER.model.createContext({
+          threads: LLM_THREADS
+        })
+        const session = new LlamaChatSession({
+          contextSequence: context.getSequence(),
+          systemPrompt: completionParams.systemPrompt
+        })
+
+        completionResult = await LLM_PROVIDER.prompt(prompt, {
+          ...completionParams,
+          session,
+          maxTokens: context.contextSize
+        })
+      } else {
+        completionResult = await LLM_PROVIDER.prompt(prompt, completionParams)
+      }
+
+      LogHelper.title(this.name)
+      LogHelper.success(`Duty executed: ${JSON.stringify(completionResult)}`)
+
+      return completionResult as unknown as LLMDutyResult
+    } catch (e) {
+      LogHelper.title(this.name)
+      LogHelper.error(`Failed to execute: ${e}`)
+    }
+
+    return null
+  }
+}
diff --git a/server/src/core/llm-manager/types.ts b/server/src/core/llm-manager/types.ts
index 28a1d8a4..8ebe607c 100644
--- a/server/src/core/llm-manager/types.ts
+++ b/server/src/core/llm-manager/types.ts
@@ -4,6 +4,7 @@ import type { LLMDuty } from '@/core/llm-manager/llm-duty'
 import type { MessageLog } from '@/types'
 
 export enum LLMDuties {
+  ActionRecognition = 'action-recognition',
   CustomNER = 'customer-ner',
   Translation = 'translation',
   Summarization = 'summarization',