feat: Add support for Langfuse LLM Tracing via OTEL (#12905)

This PR adds LLM instrumentation on langfuse for ai-editor feature

## Type of change
New feature (non-breaking change which adds functionality)

Needs langfuse account and env vars to be set

## How Has This Been Tested?

I configured personal langfuse credentials and instrumented the app,
traces can be seen in langfuse.
each conversation is one session. 
<img width="1683" height="714" alt="image"
src="https://github.com/user-attachments/assets/3fcba1c9-63cf-44b9-a355-fd6608691559"
/>
<img width="1446" height="172" alt="image"
src="https://github.com/user-attachments/assets/dfa6e98f-4741-4e04-9a9e-078d1f01e97b"
/>


## Checklist:

- [x ] My code follows the style guidelines of this project
- [ x] I have performed a self-review of my code
- [ x] I have commented on my code, particularly in hard-to-understand
areas
- [ ] I have made corresponding changes to the documentation
- [ x] My changes generate no new warnings
- [ ] I have added tests that prove my fix is effective or that my
feature works
- [ ] New and existing unit tests pass locally with my changes
- [ ] Any dependent changes have been merged and published in downstream
modules

---------

Co-authored-by: aakashb95 <aakash@chatwoot.com>
Co-authored-by: Vishnu Narayanan <iamwishnu@gmail.com>
Co-authored-by: Pranav <pranav@chatwoot.com>

lib/integrations/openai_base_service.rb

@@ -1,4 +1,6 @@
 class Integrations::OpenaiBaseService
+  include Integrations::LlmInstrumentation
+
   # gpt-4o-mini supports 128,000 tokens
   # 1 token is approx 4 characters
   # sticking with 120000 to be safe
@@ -87,21 +89,57 @@ class Integrations::OpenaiBaseService
   end
 
   def make_api_call(body)
-    headers = {
+    parsed_body = JSON.parse(body)
+    instrumentation_params = build_instrumentation_params(parsed_body)
+
+    instrument_llm_call(instrumentation_params) do
+      execute_api_request(body, parsed_body['messages'])
+    end
+  end
+
+  def build_instrumentation_params(parsed_body)
+    {
+      span_name: "llm.#{event_name}",
+      account_id: hook.account_id,
+      conversation_id: conversation&.display_id,
+      feature_name: event_name,
+      model: parsed_body['model'],
+      messages: parsed_body['messages'],
+      temperature: parsed_body['temperature']
+    }
+  end
+
+  def execute_api_request(body, messages)
+    Rails.logger.info("OpenAI API request: #{body}")
+    response = HTTParty.post(api_url, headers: api_headers, body: body)
+    Rails.logger.info("OpenAI API response: #{response.body}")
+
+    parse_api_response(response, messages)
+  end
+
+  def api_headers
+    {
       'Content-Type' => 'application/json',
       'Authorization' => "Bearer #{hook.settings['api_key']}"
     }
+  end
 
-    Rails.logger.info("OpenAI API request: #{body}")
-    response = HTTParty.post(api_url, headers: headers, body: body)
-    Rails.logger.info("OpenAI API response: #{response.body}")
+  def parse_api_response(response, messages)
+    return build_error_response(response, messages) unless response.success?
 
-    return { error: response.parsed_response, error_code: response.code } unless response.success?
+    parsed_response = JSON.parse(response.body)
+    build_success_response(parsed_response, messages)
+  end
 
-    choices = JSON.parse(response.body)['choices']
+  def build_error_response(response, messages)
+    { error: response.parsed_response, error_code: response.code, request_messages: messages }
+  end
 
-    return { message: choices.first['message']['content'] } if choices.present?
+  def build_success_response(parsed_response, messages)
+    choices = parsed_response['choices']
+    usage = parsed_response['usage']
+    message_content = choices.present? ? choices.first['message']['content'] : nil
 
-    { message: nil }
+    { message: message_content, usage: usage, request_messages: messages }
   end
 end