Migrate to WhisperX for speaker diarization

Implement a sliding window audio buffer and update the transcriber to
use WhisperX for transcription, alignment, and speaker identification.
Update the pipeline to handle and store speaker-attributed transcripts.

Additionally, update the LLM processor's reasoning parameter to
"enable_thinking".

src/llm/processor.py

@@ -83,7 +83,7 @@ class LLMProcessor:
                 model=self.model,
                 messages=messages,
                 response_format=response_format,
-                extra_body={"include_reasoning": False},
+                extra_body={"enable_thinking": False},
             )
             return response.choices[0].message.content
         except Exception as e:
@@ -125,7 +125,7 @@ class LLMProcessor:
                 model=self.model,
                 messages=messages,
                 response_format={"type": "json_object"},
-                extra_body={"include_reasoning": False},
+                extra_body={"enable_thinking": False},
             )
             logger.info("LLM Processor (Extract): Response received from backend.")