Create analysis_service.py

src/services/analysis_service.py

@@ -0,0 +1,47 @@
+import json
+import logging
+from typing import Dict, List, Any
+from crewai import Agent, Task, Crew, Process
+
+logger = logging.getLogger(__name__)
+
+class AnalysisService:
+    def __init__(self, models: Dict[str, Any]):
+        self.models = models
+        self.analyzer = models.get("deep_learning_analyzer")
+        self.rag_handler = models.get("rag_handler")
+        self.llm = models.get("llm")
+        self._create_report_agent()
+
+    def _create_report_agent(self):
+        self.report_agent = Agent(
+            role='Rédacteur de Rapports Synthétiques',
+            goal='Générer un feedback pertinent à partir du déroulement de l\'entretien',
+            backstory=(
+                "Spécialisé dans le recrutement et les ressources humaines, capable d'évaluer les candidats "
+                "sur la communication et la pertinence des réponses en fonction des questions posées, rédige "
+                "en un rapport clair, un feedback détaillé sur le candidat."
+            ),
+            allow_delegation=False,
+            verbose=False,
+            llm=self.llm
+        )
+
+    def run_analysis(self, conversation_history: List[Dict[str, Any]], job_description: str) -> Dict[str, Any]:
+        if not self.analyzer:
+            return {"error": "Analyzer non disponible"}
+
+        structured_analysis = self.analyzer.run_full_analysis(conversation_history, job_description)
+        
+        rag_feedback = []
+        if self.rag_handler:
+            rag_feedback = self._get_contextual_feedback(structured_analysis)
+        
+        report = self._generate_final_report(structured_analysis, rag_feedback)
+        
+        return report
+
+    def _get_contextual_feedback(self, structured_analysis: Dict[str, Any]) -> List[str]:
+        rag_feedback = []
+        
+        if structured_analysis.get("intent_analysis