Create app.py

app.py

@@ -0,0 +1,49 @@
+import streamlit as st
+import os
+from openai import OpenAI
+
+# Initialize the Nvidia API client using API Key stored in Streamlit secrets
+client = OpenAI(
+    base_url="https://integrate.api.nvidia.com/v1",  # Nvidia API endpoint
+    api_key=st.secrets["NVIDIA_API_KEY"]  # Nvidia API Key from Streamlit secrets
+)
+
+# Define Streamlit app layout
+st.title("AWS Well-Architected Review")
+st.write("Get recommendations for optimizing your AWS architecture.")
+
+# Session state to store messages (conversation history)
+if "messages" not in st.session_state:
+    st.session_state.messages = []
+
+# User input for AWS architecture description
+architecture_input = st.text_area("Describe your AWS architecture:")
+
+# Button to submit the input
+if st.button("Get Recommendations"):
+    if architecture_input:
+        # Add user input to the conversation
+        st.session_state.messages.append({"role": "user", "content": architecture_input})
+        
+        with st.chat_message("assistant"):
+            with st.spinner("Generating recommendations..."):
+                # Create Nvidia completion request with conversation history
+                stream = client.chat.completions.create(
+                    model="nvidia-llama-3.1-70b-instruct",  # Nvidia model name
+                    messages=st.session_state.messages,  # Include all messages in the API call
+                    temperature=0.5,
+                    top_p=0.7,
+                    max_tokens=1024,
+                    stream=True,
+                )
+
+                response_chunks = []
+                for chunk in stream:
+                    if chunk.choices[0].delta.content is not None:
+                        response_chunks.append(chunk.choices[0].delta.content)
+                response = "".join(response_chunks)
+                
+                # Display the response as recommendations
+                st.markdown(f"**Recommendations:**\n\n{response}")
+                # Add response to conversation history
+                st.session_state.messages.append({"role": "assistant", "content": response})