diff --git a/python/src/cairo_coder/core/rag_pipeline.py b/python/src/cairo_coder/core/rag_pipeline.py
index 8f27bd2..157bdfb 100644
--- a/python/src/cairo_coder/core/rag_pipeline.py
+++ b/python/src/cairo_coder/core/rag_pipeline.py
@@ -57,16 +57,6 @@ def _is_reasoning_output(self, outputs: dict[str, Any]) -> bool:
         return any(k.startswith("Thought") for k in outputs if isinstance(k, str))
 
 
-class LangsmithTracingCallback(BaseCallback):
-    @traceable()
-    def on_lm_start(self, call_id: str, instance: Any, inputs: dict[str, Any]) -> None:
-        pass
-
-    @traceable()
-    def on_lm_end(self, call_id: str, outputs: dict[str, Any], exception: Exception | None) -> None:
-        pass
-
-
 @dataclass
 class RagPipelineConfig:
     """Configuration for RAG Pipeline."""
diff --git a/python/src/cairo_coder/server/app.py b/python/src/cairo_coder/server/app.py
index 2b54e18..42b0cab 100644
--- a/python/src/cairo_coder/server/app.py
+++ b/python/src/cairo_coder/server/app.py
@@ -26,7 +26,6 @@
 from cairo_coder.core.config import VectorStoreConfig
 from cairo_coder.core.rag_pipeline import (
     AgentLoggingCallback,
-    LangsmithTracingCallback,
     RagPipeline,
 )
 from cairo_coder.core.types import Message, Role
@@ -172,7 +171,7 @@ def __init__(
         # TODO: This is the place where we should select the proper LLM configuration.
         # TODO: For now we just Hard-code DSPY - GEMINI
         dspy.configure(lm=dspy.LM("gemini/gemini-2.5-flash", max_tokens=30000), adapter=BAMLAdapter())
-        dspy.configure(callbacks=[AgentLoggingCallback(), LangsmithTracingCallback()])
+        dspy.configure(callbacks=[AgentLoggingCallback()])
         dspy.configure(track_usage=True)
 
     def _setup_routes(self):