diff --git a/README.md b/README.md
index ff1f0c5..1d23597 100644
--- a/README.md
+++ b/README.md
@@ -157,16 +157,14 @@ python run_step_function.py --reset-cache # forget previously-processed files
### Check ingestion progress
```bash
-./check_processing_status.sh
+./scripts/check_processing_status.sh
```
Reports files in DynamoDB, recently processed files, files in S3, and the remaining backlog.
### Test the chatbot
-- **Production frontend** — the CloudFront or custom-domain URL.
-- **CLI** — `python chat_test.py`
-- **Streamlit (legacy)** — `streamlit run chat_frontend.py`
+Use the production frontend — the CloudFront or custom-domain URL.
Response quality improves as more documents finish ingestion; partial answers are expected during the initial run.
diff --git a/chat_frontend.py b/chat_frontend.py
deleted file mode 100644
index ba08c1c..0000000
--- a/chat_frontend.py
+++ /dev/null
@@ -1,187 +0,0 @@
-import json
-import boto3
-import requests
-import streamlit as st
-import uuid
-import yaml
-
-config = yaml.safe_load(open("./config.yaml"))
-
-API_URL = config["rag_api_endpoint"] + "chat-response"
-FEEDBACK_URL = config["rag_api_endpoint"] + "feedback"
-# The deploy publishes the generated API key to SSM (single source of truth)
-API_KEY = boto3.client("ssm", region_name=config.get("aws_region", "us-east-1")).get_parameter(
- Name="/chatbot/api-key"
-)["Parameter"]["Value"]
-
-
-def display_response(raw_text: str):
- # Decode escaped characters like \n and \"
- decoded_text = raw_text.encode('utf-8').decode('utf-8').replace('"', "")
- st.markdown(decoded_text)
-
-
-def send_feedback_callback(timestamp: int, rating: str, feedback_text: str = ""):
- """Callback function for feedback buttons."""
- if send_feedback(st.session_state.session_id, timestamp, rating, feedback_text):
- if rating in ["thumbs_up", "thumbs_down"]:
- thumb_key = f"{timestamp}_thumb"
- st.session_state.feedback_sent.add(thumb_key)
- # Store which thumb was pressed
- rating_key = f"{timestamp}_rating"
- st.session_state[rating_key] = rating
- else:
- text_key = f"{timestamp}_text"
- st.session_state.feedback_sent.add(text_key)
- st.success("Thanks for your feedback!")
-
-
-def send_feedback(session_id: str, timestamp: int, rating: str, feedback_text: str = ""):
- """Send feedback to the API."""
- headers = {"x-api-key": API_KEY}
- data = {
- "session_id": session_id,
- "timestamp": timestamp,
- "rating": rating,
- "feedback_text": feedback_text
- }
- try:
- response = requests.post(FEEDBACK_URL, json=data, headers=headers)
- response.raise_for_status()
- return True
- except Exception as e:
- st.error(f"Error sending feedback: {e}")
- return False
-
-
-# Initialize session state
-if "messages" not in st.session_state:
- st.session_state.messages = []
-if "session_id" not in st.session_state:
- st.session_state.session_id = str(uuid.uuid4())
-if "feedback_sent" not in st.session_state:
- st.session_state.feedback_sent = set()
-
-# Streamlit App Setup
-st.set_page_config(page_title="Internet2 Chatbot PoC", page_icon="💬 ")
-st.title("Internet2 Chatbot PoC")
-
-with st.sidebar:
- st.markdown("""
- *Some questions you can ask me*
- - What workloads can I run on AWS?
- - What workloads can I run on GCP?
- - What did Lee Pang say about Amazon Omics?
- - What is AWS Omics?
- - How do I convince my leadership of the importance of FinOps practices?
- - Who has a Cloud Center of Excellence?
- - How are people doing account provisioning?
- - I've got a consultant coming in to install Control Tower for us, but they don't have any higher ed experience. What questions should I be asking to make sure I don't have to redo the work later?
- - Do I have to set up a cloud networking architecture for each platform or is there a single strategy to rule them all?
- """)
-
- # Add session reset button
- if st.button("New Conversation"):
- st.session_state.messages = []
- st.session_state.session_id = str(uuid.uuid4())
- st.session_state.feedback_sent = set()
- st.rerun()
-
-
-# Display the chat messages
-for i, msg in enumerate(st.session_state.messages):
- role = "You" if msg["role"] == "user" else "Bot"
- with st.chat_message(msg["role"]):
- display_response(msg["content"])
-
- # Add feedback buttons for assistant messages
- if msg["role"] == "assistant" and "timestamp" in msg:
- timestamp = msg["timestamp"]
- thumb_key = f"{timestamp}_thumb"
- text_key = f"{timestamp}_text"
- rating_key = f"{timestamp}_rating"
-
- col1, col2, col3 = st.columns([1, 1, 8])
-
- with col1:
- if thumb_key in st.session_state.feedback_sent:
- if st.session_state.get(rating_key) == "thumbs_up":
- st.markdown('👍
', unsafe_allow_html=True)
- else:
- st.text("👍")
- else:
- if st.button("👍", key=f"up_{i}", on_click=lambda t=timestamp: send_feedback_callback(t, "thumbs_up")):
- pass
-
- with col2:
- if thumb_key in st.session_state.feedback_sent:
- if st.session_state.get(rating_key) == "thumbs_down":
- st.markdown('👎
', unsafe_allow_html=True)
- else:
- st.text("👎")
- else:
- if st.button("👎", key=f"down_{i}", on_click=lambda t=timestamp: send_feedback_callback(t, "thumbs_down")):
- pass
-
- # Feedback text input
- if text_key not in st.session_state.feedback_sent:
- feedback_text = st.text_input("Additional feedback (optional):", key=f"feedback_{i}")
- if st.button("Submit Feedback", key=f"submit_{i}") and feedback_text:
- send_feedback_callback(timestamp, "text_feedback", feedback_text)
- else:
- st.text("✓ Text feedback submitted")
-
-
-# User input field
-user_input = st.chat_input("Type your question here...")
-
-if user_input:
- # Add user message to history
- st.session_state.messages.append({"role": "user", "content": user_input})
- with st.chat_message("user"):
- display_response(user_input)
-
- # Send to API with session ID
- headers = {"x-api-key": API_KEY}
- data = {
- "query": user_input,
- "session_id": st.session_state.session_id
- }
- try:
- response = requests.post(API_URL, json=data, headers=headers)
- response.raise_for_status()
- response_data = json.loads(response.text)
- bot_reply = response_data.get("response", response.text)
- timestamp = response_data.get("timestamp")
- # Update session ID if provided
- if "session_id" in response_data:
- st.session_state.session_id = response_data["session_id"]
- except Exception as e:
- bot_reply = f"Error: {e}"
- timestamp = None
-
- # Add bot response to history with timestamp
- message_data = {"role": "assistant", "content": bot_reply}
- if timestamp:
- message_data["timestamp"] = timestamp
-
- st.session_state.messages.append(message_data)
- with st.chat_message("assistant"):
- display_response(bot_reply)
-
- # Add feedback buttons for the new response
- if timestamp:
- col1, col2, col3 = st.columns([1, 1, 8])
-
- with col1:
- if st.button("👍", key=f"up_new", on_click=lambda: send_feedback_callback(timestamp, "thumbs_up")):
- pass
-
- with col2:
- if st.button("👎", key=f"down_new", on_click=lambda: send_feedback_callback(timestamp, "thumbs_down")):
- pass
-
- # Feedback text input
- feedback_text = st.text_input("Additional feedback (optional):", key=f"feedback_new")
- if st.button("Submit Feedback", key=f"submit_new") and feedback_text:
- send_feedback_callback(timestamp, "text_feedback", feedback_text)
diff --git a/chat_test.py b/chat_test.py
deleted file mode 100644
index f2da4ce..0000000
--- a/chat_test.py
+++ /dev/null
@@ -1,80 +0,0 @@
-import json
-import boto3
-import requests
-import uuid
-import yaml
-
-config = yaml.safe_load(open("./config.yaml"))
-
-API_URL = config["rag_api_endpoint"] + "chat-response"
-FEEDBACK_URL = config["rag_api_endpoint"] + "feedback"
-# The deploy publishes the generated API key to SSM (single source of truth)
-API_KEY = boto3.client("ssm", region_name=config.get("aws_region", "us-east-1")).get_parameter(
- Name="/chatbot/api-key"
-)["Parameter"]["Value"]
-
-headers = {
- "x-api-key": API_KEY,
-}
-
-session_id = str(uuid.uuid4())
-
-def format_response(raw_text: str):
- # Decode escaped characters like \n and \"
- decoded_text = raw_text.encode('utf-8').decode('utf-8').replace('"', "")
- return decoded_text
-
-def send_feedback(session_id: str, timestamp: int, rating: str, feedback_text: str = ""):
- """Send feedback to the API."""
- data = {
- "session_id": session_id,
- "timestamp": timestamp,
- "rating": rating,
- "feedback_text": feedback_text
- }
- try:
- response = requests.post(FEEDBACK_URL, json=data, headers=headers)
- response.raise_for_status()
- return True
- except Exception as e:
- print(f"Error sending feedback: {e}")
- return False
-
-print("Internet2 Chatbot - Type 'quit' to exit")
-print(f"Session ID: {session_id}")
-print("-" * 50)
-
-while True:
- question = input("\nAsk a question: ")
- if question.lower() == 'quit':
- break
-
- data = {
- "query": question,
- "session_id": session_id
- }
-
- try:
- response = requests.post(API_URL, json=data, headers=headers)
- response.raise_for_status()
- response_data = json.loads(response.text)
- bot_reply = response_data.get("response", response.text)
- timestamp = response_data.get("timestamp")
- print(f"\nBot: {format_response(bot_reply)}")
-
- # Ask for feedback
- if timestamp:
- feedback = input("\nRate this response (u=👍, d=👎, enter=skip): ").lower()
- if feedback == 'u':
- if send_feedback(session_id, timestamp, "thumbs_up"):
- print("Thanks for your feedback!")
- elif feedback == 'd':
- if send_feedback(session_id, timestamp, "thumbs_down"):
- print("Thanks for your feedback!")
- # Ask for additional feedback on thumbs down
- text_feedback = input("Any additional feedback? (optional): ")
- if text_feedback:
- send_feedback(session_id, timestamp, "text_feedback", text_feedback)
-
- except Exception as e:
- print(f"Error: {e}")
diff --git a/requirements.txt b/requirements.txt
index ce68398..a2b7ef6 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -3,7 +3,6 @@ boto3==1.42.84
requests-aws4auth==1.3.1
pyyaml==6.0.3
requests==2.33.1
-streamlit==1.56.0
aws-cdk-lib==2.248.0
constructs>=10.0.0,<11.0.0
beautifulsoup4==4.14.3
diff --git a/check_processing_status.sh b/scripts/check_processing_status.sh
similarity index 100%
rename from check_processing_status.sh
rename to scripts/check_processing_status.sh