Implement batch processing for LLM-based tax analysis and enhance match confidence scoring

app/services/matching_engine.py

@@ -25,11 +25,36 @@ class MatchingEngine:
             receipts, transactions
         )
 
-        # Apply rules and enhance matches
-        enhanced_matches = []
+        # Apply traditional rules first (lightweight, no API calls)
         for match in ai_matches:
-            enhanced_match = self._enhance_match_with_rules(match, user_location)
-            enhanced_matches.append(enhanced_match)
+            rule_results = self.rules_engine.apply_rules(
+                match.receipt, match.transaction
+            )
+
+            # Apply confidence boost from traditional rules
+            if rule_results["confidence_boost"] > 0:
+                match.confidence_score = min(
+                    1.0, match.confidence_score + rule_results["confidence_boost"]
+                )
+
+            # Auto-approve if rules say so
+            if rule_results["auto_approve"]:
+                match.confidence_score = 1.0
+                match.match_reason += " (Auto-approved by rules)"
+
+        # Now apply LLM-based tax analysis in a SINGLE batch call
+        try:
+            enhanced_matches = self.llm_tax_analyzer.analyze_and_apply_tax_rules_batch(
+                ai_matches, user_location
+            )
+        except Exception as e:
+            # If batch LLM analysis fails, log it and continue with matches as-is
+            import logging
+
+            logging.error(f"Batch LLM tax analysis failed: {str(e)}")
+            for match in ai_matches:
+                match.match_reason += " (Note: Advanced tax analysis unavailable)"
+            enhanced_matches = ai_matches
 
         return enhanced_matches