added streaming response and conversation memory. Also added reranker, ragas. Added 1 test

app/core/llm.py

@@ -41,3 +41,18 @@ Answer:""",
             "question": question,
             "context": context
         })
+
+async def generate_stream(self, question: str, context: str):
+    """Stream LLM responses"""
+    async for chunk in self.chain.astream({
+        "question": question,
+        "context": context
+    }):
+        yield chunk
+
+from langchain.memory import ConversationBufferMemory
+
+class ConversationalRAG:
+    def __init__(self, rag_chain):
+        self.rag_chain = rag_chain
+        self.memory = ConversationBufferMemory()

pyproject.toml

@@ -14,6 +14,8 @@ dependencies = [
     "python-docx>=1.2.0",
     "python-multipart>=0.0.20",
     "qdrant-client>=1.15.1",
+    "ragas>=0.3.7",
+    "rank-bm25>=0.2.2",
     "sentence-transformers>=5.1.2",
     "unstructured>=0.18.15",
     "uvicorn>=0.38.0",

tests/test_rag.py

@@ -0,0 +1,10 @@
+import pytest
+from app.core.embeddings import EmbeddingGenerator
+
+def test_embedding_generation():
+    embedder = EmbeddingGenerator("sentence-transformers/all-MiniLM-L6-v2")
+    embeddings = embedder.generate(["test text"])
+    assert len(embeddings) == 1
+    assert len(embeddings[0]) == 384
+
+# Run tests with: uv run pytest