Cloud & AI SolutionsReimagined

import boto3
import pandas as pd
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split

# Connect to AWS S3
s3 = boto3.client('s3')
print("🔌 Connected to AWS S3")

# Download dataset from S3
bucket_name = 'ai-analytics-data'
file_key = 'customer_data.parquet'
local_file = '/tmp/customer_data.parquet'

s3.download_file(bucket_name, file_key, local_file)
print(f"📥 Downloaded {file_key} from S3")

# Load and prepare data
df = pd.read_parquet(local_file)
print(f"✅ Loaded dataset with {len(df)} records")

# Prepare features and target
X = df.drop('conversion', axis=1)
y = df['conversion']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

# Train ML model
model = RandomForestClassifier(n_estimators=100)
model.fit(X_train, y_train)
accuracy = model.score(X_test, y_test)
print(f"🧠 Model trained with {accuracy:.2%} accuracy")

# Save predictions back to S3
predictions = model.predict(X_test)
results_df = pd.DataFrame({
    'customer_id': X_test.index,
    'prediction': predictions
})
results_df.to_parquet('/tmp/predictions.parquet')

s3.upload_file(
    '/tmp/predictions.parquet', 
    bucket_name, 
    'predictions/customer_predictions.parquet'
)
print("🚀 Uploaded predictions to S3")