RAG1

agents/__init__.py

@@ -0,0 +1,2 @@
+from .rag_decision_agent import rag_decision
+from .rag_retrieval_agent import get_top_k

agents/rag_decision_agent.py

@@ -0,0 +1,42 @@
+from openai import OpenAI
+import os
+import re
+
+
+def rag_decision(query):
+    """
+    Function to perform whether RAG is required for this user's query.
+    """
+    client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
+    prompt = f"""
+    You are an agent of a Diagnostics Lab agentic ai Chatbot system. 
+    Your task is to determine whether the user's query requires Retrieval-Augmented 
+    Generation (RAG) or not. Currently the system has access to the rates and offers
+    of the services provided by the diagnostics lab only.
+    If the query or part of the query is about the rates, discounts and offers 
+    respond with "RAG NEEDED" and nothing else
+    If the query is general and can be answered with your knowledge respond with 
+    "RAG NOT NEEDED" and nothing else.
+
+    User's query: {query}
+    """
+    # Make the API call to OpenAI
+    response = client.chat.completions.create(
+        model="gpt-4o-mini",
+        messages=[
+            {
+                "role": "user",
+                "content": [
+                    {
+                        "type": "text",
+                        "text": prompt,
+                    },
+                ],
+            }
+        ],
+    )   
+    decision = response.choices[0].message.content.strip()
+    if re.search(r'\brag needed\b', decision, re.IGNORECASE):
+        return True
+    else:
+        return False

app.py

@@ -4,11 +4,14 @@ import os
 import base64
 import time
 import copy
-from rag import get_top_k
+
 from dotenv import load_dotenv
 # Load environment variables from .env file
 load_dotenv()
 
+from agents import rag_decision
+from agents import get_top_k
+
 client = OpenAI(api_key=os.getenv("OPENAI_API_KEY"))
 
 
@@ -23,10 +26,13 @@ with gr.Blocks() as demo:
     )
     clear = gr.ClearButton([chat_input, chatbot])
 
-    def use_rag(query, k=3):
-        """
-        Use RAG to get top k results for the query.
-        """
+    # Agent1 - RAG Decision Agent (whether RAG is needed for the user's query)
+    def agent1_rag_decision(query):
+        decision = rag_decision(query)
+        return decision
+    
+    # Agent2 - RAG Retrieval Agent (retrieve top k relevant documents)
+    def agent2_use_rag(query, k=3):
         results = get_top_k(query, k=k)
         return results
     

rag/__init__.py

@@ -1 +0,0 @@
-from .retriever import get_top_k