Memory Chatbot
Defining parameters.
This section defines key parameters for the notebook. Parameters control the behavior of the code, making it easy to customize without altering the logic. By centralizing parameters at the start, we ensure better readability, maintainability, and adaptability for different use cases.
app_name = "memory-chatbot" # E.g. 'api-chatbot'
deployment_setting_key = "appdepl"
app_prefix = "apps"
app_dir = None
assert app_name, "Please enter application name"
assert deployment_setting_key, "Please enter deployment_setting_key"
assert app_prefix, "Please enter app_prefix"
Preparing Data
import os
import requests
# Create the GitHub API URL
url = "https://api.github.com/repos/practicusai/sample-data/contents/ecomm?ref=main"
# Call the API
response = requests.get(url)
if response.status_code == 200:
files = response.json() # Get response in JSON format
for file in files:
file_url = file["download_url"]
file_name = file["name"]
# Download files
file_response = requests.get(file_url)
if file_response.status_code == 200:
with open(file_name, "wb") as f:
f.write(file_response.content)
print(f"'{file_name}' successfully downloaded.")
else:
print(f"'{file_name}' file failed to download.")
else:
print(f"Failed to retrieve data from API, HTTP status: {response.status_code}")
After testing our application we can set our configurations and start the deployment process.
If you don't know your prefixes and deployments you can check them out by using the SDK like down below:
my_app_list = prt.apps.get_list()
display(my_app_list.to_pandas())
print("Using first app name:", app_name)
my_app_prefix_list = prt.apps.get_prefix_list()
display(my_app_prefix_list.to_pandas())
print("Using first app prefix", app_prefix)
my_app_settings = prt.apps.get_deployment_setting_list()
display(my_app_settings.to_pandas())
print("Using first setting with key:", deployment_setting_key)
Deploying app
prt.apps.deploy(
deployment_setting_key=deployment_setting_key, # Deployment Key, ask admin for deployment key
prefix=app_prefix, # Apphost deployment extension
app_name=app_name,
app_dir=None, # Directory of files that will be deployed ('None' for current directory)
)
Supplementary Files
Home.py
import streamlit as st
import pandas as pd
import practicuscore as prt
from practicuscore.gen_ai import PrtLangMessage, PrtLangRequest
import requests
st.set_page_config(page_title="E-Commerce Product Review & Chatbot", layout="wide")
def load_data():
excel_file = "ecommerce_data_with_images.xlsx"
products_df = pd.read_excel(excel_file, sheet_name="Products")
reviews_df = pd.read_excel(excel_file, sheet_name="Reviews")
sizes_df = pd.read_excel(excel_file, sheet_name="Sizes")
images_df = pd.read_excel(excel_file, sheet_name="ProductImages")
return products_df, reviews_df, sizes_df, images_df
products_df, reviews_df, sizes_df, images_df = load_data()
st.title("🛍️ E-Commerce Product Review & Chatbot")
col1, col2 = st.columns([3, 1])
with col1:
selected_product = st.selectbox("Select a product:", products_df["product_name"].unique())
with col2:
show_image = st.button("Show Image")
if show_image and selected_product:
product_id = products_df.loc[products_df["product_name"] == selected_product, "product_id"].values[0]
image_url = images_df.loc[images_df["product_id"] == product_id, "image_url"].values[0]
st.image(image_url, caption=selected_product, width=300)
selected_rating = st.slider("Filter reviews by rating:", min_value=1, max_value=5, value=(1, 5))
if selected_product:
product_id = products_df.loc[products_df["product_name"] == selected_product, "product_id"].values[0]
product_reviews = reviews_df[
(reviews_df["product_id"] == product_id)
& (reviews_df["rating"].between(selected_rating[0], selected_rating[1]))
]
available_sizes = sizes_df["size"].unique()
st.subheader("Customer Reviews")
review_columns = st.columns(2)
for i, (_, row) in enumerate(product_reviews.iterrows()):
with review_columns[i % 2]:
st.markdown(
f"""
<div style='padding: 10px; border-radius: 10px; background-color: #2c3e50; color: #ecf0f1; margin-bottom: 10px;'>
<strong>⭐ {row["rating"]}/5</strong><br>
{row["review"]}
</div>
""",
unsafe_allow_html=True,
)
st.subheader("Available Sizes")
st.write(", ".join(available_sizes))
import time
def analyze_sentiment(reviews):
if reviews.empty:
return ["No reviews available."]
api_url = f"https://dev.practicus.io/models/llm-proxy/"
token = prt.models.get_session_token(api_url=api_url)
context = f"You are an expert product assistant providing details about e-commerce products based on available data.\nReviews: {', '.join(reviews['review'].tolist())}"
practicus_llm_req = PrtLangRequest(
messages=[PrtLangMessage(content=context, role="human")], lang_model="model", streaming=True
)
headers = {"authorization": f"Bearer {token}", "content-type": "application/json"}
data_js = practicus_llm_req.model_dump_json(indent=2, exclude_unset=True)
results = []
placeholder = st.empty()
with requests.post(api_url, headers=headers, data=data_js, stream=True) as r:
for word in r.iter_content(1024):
word_decoded = word.decode("utf-8")
results.append(word_decoded)
placeholder.markdown("".join(results))
time.sleep(0.1)
return "".join(results)
def analyze_sentiment2(reviews):
if reviews.empty:
return ["No reviews available."]
api_url = f"https://dev.practicus.io/models/llm-proxy/"
token = prt.models.get_session_token(api_url=api_url)
context = f"Based on the response above, which specific parts of the reviews contributed to the answer? Please extract relevant quotes.\nReviews: {', '.join(product_reviews['review'].tolist()[:10])}"
practicus_llm_req = PrtLangRequest(
messages=[PrtLangMessage(content=context, role="human")], lang_model="model", streaming=True
)
headers = {"authorization": f"Bearer {token}", "content-type": "application/json"}
data_js = practicus_llm_req.model_dump_json(indent=2, exclude_unset=True)
results = []
placeholder = st.empty()
with requests.post(api_url, headers=headers, data=data_js, stream=True) as r:
for word in r.iter_content(1024):
word_decoded = word.decode("utf-8")
results.append(word_decoded)
placeholder.markdown("".join(results))
time.sleep(0.1)
return "".join(results)
st.sidebar.title("🗨️ Chat with AI Product Assistant")
user_input = st.sidebar.text_input("Ask about the product:")
if user_input:
sentiment_summary = analyze_sentiment(product_reviews)
st.sidebar.write(sentiment_summary)
st.sidebar.subheader("🔍 Relevant Context from Reviews")
sentiment_summary2 = analyze_sentiment2(product_reviews)
st.sidebar.write(sentiment_summary2)
st.sidebar.subheader("💡 Suggested Questions")
if selected_product:
context = f"Product: {selected_product}\nReviews: {', '.join(product_reviews['review'].tolist())}"
suggested_questions = [
"What is the overall customer satisfaction for this product?",
"Are there any common complaints about this product?",
"What do most customers like about this product?",
]
for question in suggested_questions:
st.sidebar.write(f"- {question}")
Previous: Langflow API | Next: Cv Assistant > Cv Assistant