--- a
+++ b/breast-cancer-rag-app/backend/main.py
@@ -0,0 +1,146 @@
+import os
+import json
+from fastapi import FastAPI, HTTPException
+from pydantic import BaseModel
+from azure.identity import DefaultAzureCredential
+from azure.kusto.data import KustoClient, KustoConnectionStringBuilder
+from openai import AzureOpenAI
+from tenacity import retry, wait_random_exponential, stop_after_attempt
+from fastapi.middleware.cors import CORSMiddleware
+
+app = FastAPI()
+
+# Add CORS middleware
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+# Configuration
+OPENAI_GPT4_DEPLOYMENT = os.getenv("OPENAI_GPT4_DEPLOYMENT")
+OPENAI_ENDPOINT = os.getenv("OPENAI_ENDPOINT")
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+OPENAI_ADA_DEPLOYMENT = os.getenv("OPENAI_ADA_DEPLOYMENT")
+
+KUSTO_URI = os.getenv("KUSTO_URI")
+KUSTO_DATABASE = os.getenv("KUSTO_DATABASE")
+KUSTO_TABLE = os.getenv("KUSTO_TABLE")
+
+# Azure Authentication
+credential = DefaultAzureCredential()
+
+# # Kusto Client
+# kcsb = KustoConnectionStringBuilder.with_aad_application_key_authentication(
+#     KUSTO_URI,
+#     application_client_id=os.getenv("AZURE_CLIENT_ID"),
+#     application_key=os.getenv("AZURE_CLIENT_SECRET"),
+#     authority_id=os.getenv("AZURE_TENANT_ID")
+# )
+
+# Use this universal approach that works with most versions
+kcsb = KustoConnectionStringBuilder.with_aad_application_key_authentication(
+    KUSTO_URI,
+    os.getenv("AZURE_CLIENT_ID"),
+    os.getenv("AZURE_CLIENT_SECRET"),
+    os.getenv("AZURE_TENANT_ID")
+)
+
+kusto_client = KustoClient(kcsb)
+
+# OpenAI Client
+openai_client = AzureOpenAI(
+    azure_endpoint=OPENAI_ENDPOINT,
+    api_key=OPENAI_API_KEY,
+    api_version="2023-09-01-preview"
+)
+
+@retry(wait=wait_random_exponential(min=1, max=20), stop=stop_after_attempt(6))
+def generate_embeddings(text: str) -> list:
+    try:
+        response = openai_client.embeddings.create(
+            input=[text.replace("\n", " ")],
+            model=OPENAI_ADA_DEPLOYMENT
+        )
+        return response.data[0].embedding
+    except Exception as e:
+        print(f"Embedding generation failed: {str(e)}")
+        return None
+
+def execute_kusto_query(query: str) -> list:
+    try:
+        response = kusto_client.execute(KUSTO_DATABASE, query)
+        return [row.to_dict() for row in response.primary_results[0]]
+    except Exception as e:
+        print(f"Kusto query failed: {str(e)}")
+        return []
+
+def query_biospecimen_data(question: str, top_results: int = 3) -> dict:
+    embedding = generate_embeddings(question)
+    if not embedding:
+        return {"answer": "Embedding generation failed", "sources": []}
+
+    embedding_str = "[" + ",".join(map(str, embedding)) + "]"
+    kusto_query = f"""
+    {KUSTO_TABLE}
+    | extend similarity = cosine_similarity(embedding, dynamic({embedding_str}))
+    | top {top_results} by similarity desc
+    | project content, metadata, similarity
+    """
+    
+    results = execute_kusto_query(kusto_query)
+    if not results:
+        return {"answer": "No relevant records found", "sources": []}
+
+    context = "\n".join([
+        f"Record {idx+1} (Similarity: {row['similarity']:.2f}):\n"
+        f"Content: {row['content']}\n"
+        f"Metadata: {json.dumps(row['metadata'])}\n"
+        for idx, row in enumerate(results)
+    ])
+
+    try:
+        response = openai_client.chat.completions.create(
+            model=OPENAI_GPT4_DEPLOYMENT,
+            messages=[
+                {"role": "system", "content": "You are a biomedical research assistant. "
+                 "Use only the provided context to answer questions about breast cancer biospecimens."},
+                {"role": "user", "content": f"Question: {question}\nContext:\n{context}"}
+            ],
+            temperature=0.2,
+            max_tokens=500
+        )
+        answer = response.choices[0].message.content
+    except Exception as e:
+        answer = f"LLM processing error: {str(e)}"
+
+    return {
+        "answer": answer,
+        "sources": [{
+            "content": r["content"],
+            "metadata": r["metadata"],
+            "similarity": round(r["similarity"], 2)
+        } for r in results],
+        "processing_time": f"{len(results)} results analyzed"
+    }
+
+class QueryRequest(BaseModel):
+    question: str
+    top_results: int = 3
+
+@app.post("/query")
+async def handle_query(request: QueryRequest):
+    try:
+        result = query_biospecimen_data(request.question, request.top_results)
+        return result
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@app.get("/health")
+async def health_check():
+    return {"status": "healthy", "service": "biospecimen-query"}
+
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=int(os.getenv("PORT", 8000)))
\ No newline at end of file