email marketing

Using Streamlit
1

import pandas as pd
import numpy as np
import streamlit as st
from sklearn.cluster import KMeans
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import train_test_split

Streamlit configuration

st.set_page_config(page_title=“AI Email Marketing Platform”, page_icon=“:e_mail:”, layout=“wide”)
st.title(“:open_mailbox_with_raised_flag: AI-Driven Email Marketing Platform”)
st.markdown(“Upload customer data to analyze engagement, segment users, and generate smart email content.”)

File uploader

file = st.file_uploader(“Upload CSV file with customer data”, type=[“csv”])

if file:
df = pd.read_csv(file)
st.success(“:white_check_mark: File uploaded successfully!”)
st.write(“### :magnifying_glass_tilted_left: Data Preview:”)
st.dataframe(df.head())

# Check if required columns are present
if 'Email Opened' in df.columns and 'Discount Offered' in df.columns:
    # Preprocessing
    features = ['Discount Offered']
    X = df[features]
    y = df['Email Opened']

    # Train/test split
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

    # Train model
    model = LogisticRegression()
    model.fit(X_train, y_train)

    # Predict engagement probabilities
    df['Engagement Probability'] = model.predict_proba(X)[:, 1]
    st.subheader("📈 Predicted Engagement Probabilities")
    st.dataframe(df[['Customer ID', 'Discount Offered', 'Email Opened', 'Engagement Probability']].head())

    # Clustering users based on engagement
    st.subheader("👥 Customer Segmentation (KMeans)")
    kmeans = KMeans(n_clusters=3, random_state=42)
    df['Segment'] = kmeans.fit_predict(df[['Engagement Probability']])
    st.write("### 📊 Segment Distribution")
    st.write(df['Segment'].value_counts().rename("Users").reset_index(names=["Segment"]))

    # Download segmented data
    st.download_button(
        label="⬇️ Download Segmented Data",
        data=df.to_csv(index=False).encode("utf-8"),
        file_name="segmented_customers.csv",
        mime="text/csv"
    )

    # Generate dummy personalized messages
    st.subheader("✉️ Sample Personalized Emails")
    for i in range(3):
        segment_df = df[df['Segment'] == i]
        sample_msg = (
            f"Hi! As a valued customer, we’ve curated a special offer for you "
            f"based on your past interest. Check it out now!"
        )
        st.markdown(f"**Segment {i} ({len(segment_df)} users)**: {sample_msg}")
else:
    st.warning("⚠️ Required columns 'Email Opened' and 'Discount Offered' not found in CSV.")

else:
st.info(“:open_file_folder: Please upload a CSV file to begin.”)