fix(llm): resolve score runtime config from saved profiles

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

rag_eval/metrics/factory.py

@@ -27,13 +27,48 @@ from ragas.metrics.collections import (
 from .pipeline import MetricPipeline
 
 
+def _resolve_openai_client_kwargs(
+    judge_model: str,
+    settings: EvaluationSettings,
+) -> dict[str, Any]:
+    """Return AsyncOpenAI kwargs, preferring a matching LLM Profile over .env settings.
+
+    Lookup order:
+      1. LLM Profile whose model name equals judge_model (exact match)
+      2. Fall back to EvaluationSettings (.env)
+    """
+    try:
+        # Lazy import to avoid circular dependency (webapp -> rag_eval is one-way).
+        from webapp.services.profile_manager import profile_manager
+        profiles = profile_manager.list_all()
+        for profile in profiles:
+            if profile.model == judge_model:
+                kwargs: dict[str, Any] = {
+                    "api_key": profile.api_key or "sk-placeholder",
+                    "timeout": float(profile.timeout_seconds or 30),
+                }
+                if profile.base_url and profile.base_url.strip():
+                    kwargs["base_url"] = profile.base_url.strip()
+                return kwargs
+    except Exception:  # noqa: BLE001
+        # If profile lookup fails for any reason, fall through to .env settings.
+        pass
+
+    return settings.openai_client_kwargs
+
+
 def build_models(
     judge_model: str,
     embedding_model: str,
     settings: EvaluationSettings,
 ) -> tuple[Any, Any]:
-    """Create the LLM and embedding clients required by the selected RAGAS metrics."""
-    client = AsyncOpenAI(**settings.openai_client_kwargs)
+    """Create the LLM and embedding clients required by the selected RAGAS metrics.
+
+    Dynamically resolves connection settings from the stored LLM Profiles first
+    (matched by model name), falling back to .env settings when no profile matches.
+    """
+    client_kwargs = _resolve_openai_client_kwargs(judge_model, settings)
+    client = AsyncOpenAI(**client_kwargs)
     llm = llm_factory(judge_model, client=client)
     embeddings = embedding_factory(provider="openai", model=embedding_model, client=client)
     return llm, embeddings