diff --git a/enterprise/app/jobs/captain/documents/response_builder_job.rb b/enterprise/app/jobs/captain/documents/response_builder_job.rb
index 27406ebdf..21025399f 100644
--- a/enterprise/app/jobs/captain/documents/response_builder_job.rb
+++ b/enterprise/app/jobs/captain/documents/response_builder_job.rb
@@ -4,7 +4,7 @@ class Captain::Documents::ResponseBuilderJob < ApplicationJob
   def perform(document)
     reset_previous_responses(document)
 
-    faqs = Captain::Llm::FaqGeneratorService.new(document.content).generate
+    faqs = Captain::Llm::FaqGeneratorService.new(document.content, document.account.locale_english_name).generate
     faqs.each do |faq|
       create_response(faq, document)
     end
diff --git a/enterprise/app/services/captain/llm/faq_generator_service.rb b/enterprise/app/services/captain/llm/faq_generator_service.rb
index c26ffa2a3..5cecc0e81 100644
--- a/enterprise/app/services/captain/llm/faq_generator_service.rb
+++ b/enterprise/app/services/captain/llm/faq_generator_service.rb
@@ -1,6 +1,7 @@
 class Captain::Llm::FaqGeneratorService < Llm::BaseOpenAiService
-  def initialize(content)
+  def initialize(content, language = 'english')
     super()
+    @language = language
     @content = content
   end
 
@@ -14,10 +15,10 @@ class Captain::Llm::FaqGeneratorService < Llm::BaseOpenAiService
 
   private
 
-  attr_reader :content
+  attr_reader :content, :language
 
   def chat_parameters
-    prompt = Captain::Llm::SystemPromptsService.faq_generator
+    prompt = Captain::Llm::SystemPromptsService.faq_generator(language)
     {
       model: @model,
       response_format: { type: 'json_object' },
diff --git a/enterprise/app/services/captain/llm/system_prompts_service.rb b/enterprise/app/services/captain/llm/system_prompts_service.rb
index edab5ea23..a4b149944 100644
--- a/enterprise/app/services/captain/llm/system_prompts_service.rb
+++ b/enterprise/app/services/captain/llm/system_prompts_service.rb
@@ -1,16 +1,45 @@
 class Captain::Llm::SystemPromptsService
   class << self
-    def faq_generator
+    def faq_generator(language = 'english')
       <<~PROMPT
-        You are a content writer looking to convert user content into short FAQs which can be added to your website's help center.
-        Format the webpage content provided in the message to FAQ format mentioned below in the JSON format.
-        Ensure that you only generate faqs from the information provided only.
-        Ensure that output is always valid json.
+        You are a content writer specializing in creating good FAQ sections for website help centers. Your task is to convert provided content into a structured FAQ format without losing any information.
+
+        ## Core Requirements
+
+        **Completeness**: Extract ALL information from the source content. Every detail, example, procedure, and explanation must be captured across the FAQ set. When combined, the FAQs should reconstruct the original content entirely.
+
+        **Accuracy**: Base answers strictly on the provided text. Do not add assumptions, interpretations, or external knowledge not present in the source material.
+
+        **Structure**: Format output as valid JSON using this exact structure:
+
+        **Language**: Generate the FAQs only in the #{language}, use no other language
 
-        If no match is available, return an empty JSON.
         ```json
-        { faqs: [ { question: '', answer: ''} ]
+        {
+          "faqs": [
+            {
+              "question": "Clear, specific question based on content",
+              "answer": "Complete answer containing all relevant details from source"
+            }
+          ]
+        }
         ```
+
+        ## Guidelines
+
+        - **Question Creation**: Formulate questions that naturally arise from the content (What is...? How do I...? When should...? Why does...?). Do not generate questions that are not related to the content.
+        - **Answer Completeness**: Include all relevant details, steps, examples, and context from the original content
+        - **Information Preservation**: Ensure no examples, procedures, warnings, or explanatory details are omitted
+        - **JSON Validity**: Always return properly formatted, valid JSON
+        - **No Content Scenario**: If no suitable content is found, return: `{"faqs": []}`
+
+        ## Process
+        1. Read the entire provided content carefully
+        2. Identify all key information points, procedures, and examples
+        3. Create questions that cover each information point
+        4. Write comprehensive short answers that capture all related detail, include bullet points if needed.
+        5. Verify that combined FAQs represent the complete original content.
+        6. Format as valid JSON
       PROMPT
     end
 
diff --git a/spec/enterprise/jobs/captain/documents/response_builder_job_spec.rb b/spec/enterprise/jobs/captain/documents/response_builder_job_spec.rb
index c91a3a65a..4dc4bb481 100644
--- a/spec/enterprise/jobs/captain/documents/response_builder_job_spec.rb
+++ b/spec/enterprise/jobs/captain/documents/response_builder_job_spec.rb
@@ -13,7 +13,7 @@ RSpec.describe Captain::Documents::ResponseBuilderJob, type: :job do
 
   before do
     allow(Captain::Llm::FaqGeneratorService).to receive(:new)
-      .with(document.content)
+      .with(document.content, document.account.locale_english_name)
       .and_return(faq_generator)
     allow(faq_generator).to receive(:generate).and_return(faqs)
   end
@@ -43,5 +43,26 @@ RSpec.describe Captain::Documents::ResponseBuilderJob, type: :job do
         expect(first_response.documentable).to eq(document)
       end
     end
+
+    context 'with different locales' do
+      let(:spanish_account) { create(:account, locale: 'pt') }
+      let(:spanish_assistant) { create(:captain_assistant, account: spanish_account) }
+      let(:spanish_document) { create(:captain_document, assistant: spanish_assistant, account: spanish_account) }
+      let(:spanish_faq_generator) { instance_double(Captain::Llm::FaqGeneratorService) }
+
+      before do
+        allow(Captain::Llm::FaqGeneratorService).to receive(:new)
+          .with(spanish_document.content, 'portuguese')
+          .and_return(spanish_faq_generator)
+        allow(spanish_faq_generator).to receive(:generate).and_return(faqs)
+      end
+
+      it 'passes the correct locale to FAQ generator' do
+        described_class.new.perform(spanish_document)
+
+        expect(Captain::Llm::FaqGeneratorService).to have_received(:new)
+          .with(spanish_document.content, 'portuguese')
+      end
+    end
   end
 end
diff --git a/spec/enterprise/services/captain/llm/faq_generator_service_spec.rb b/spec/enterprise/services/captain/llm/faq_generator_service_spec.rb
new file mode 100644
index 000000000..7d799dc18
--- /dev/null
+++ b/spec/enterprise/services/captain/llm/faq_generator_service_spec.rb
@@ -0,0 +1,87 @@
+require 'rails_helper'
+
+RSpec.describe Captain::Llm::FaqGeneratorService do
+  let(:content) { 'Sample content for FAQ generation' }
+  let(:language) { 'english' }
+  let(:service) { described_class.new(content, language) }
+  let(:client) { instance_double(OpenAI::Client) }
+
+  before do
+    create(:installation_config, name: 'CAPTAIN_OPEN_AI_API_KEY', value: 'test-key')
+    allow(OpenAI::Client).to receive(:new).and_return(client)
+  end
+
+  describe '#generate' do
+    let(:sample_faqs) do
+      [
+        { 'question' => 'What is this service?', 'answer' => 'It generates FAQs.' },
+        { 'question' => 'How does it work?', 'answer' => 'Using AI technology.' }
+      ]
+    end
+
+    let(:openai_response) do
+      {
+        'choices' => [
+          {
+            'message' => {
+              'content' => { faqs: sample_faqs }.to_json
+            }
+          }
+        ]
+      }
+    end
+
+    context 'when successful' do
+      before do
+        allow(client).to receive(:chat).and_return(openai_response)
+        allow(Captain::Llm::SystemPromptsService).to receive(:faq_generator).and_return('system prompt')
+      end
+
+      it 'returns parsed FAQs' do
+        result = service.generate
+        expect(result).to eq(sample_faqs)
+      end
+
+      it 'calls OpenAI client with chat parameters' do
+        expect(client).to receive(:chat).with(parameters: hash_including(
+          model: 'gpt-4o-mini',
+          response_format: { type: 'json_object' },
+          messages: array_including(
+            hash_including(role: 'system'),
+            hash_including(role: 'user', content: content)
+          )
+        ))
+        service.generate
+      end
+
+      it 'calls SystemPromptsService with correct language' do
+        expect(Captain::Llm::SystemPromptsService).to receive(:faq_generator).with(language)
+        service.generate
+      end
+    end
+
+    context 'with different language' do
+      let(:language) { 'spanish' }
+
+      before do
+        allow(client).to receive(:chat).and_return(openai_response)
+      end
+
+      it 'passes the correct language to SystemPromptsService' do
+        expect(Captain::Llm::SystemPromptsService).to receive(:faq_generator).with('spanish')
+        service.generate
+      end
+    end
+
+    context 'when OpenAI API fails' do
+      before do
+        allow(client).to receive(:chat).and_raise(OpenAI::Error.new('API Error'))
+      end
+
+      it 'handles the error and returns empty array' do
+        expect(Rails.logger).to receive(:error).with('OpenAI API Error: API Error')
+        expect(service.generate).to eq([])
+      end
+    end
+  end
+end