Customer sentiment analysis, which evaluates attitudes and opinions expressed in product reviews, has become crucial in e-commerce, with textual data providing valuable unstructured information. Despite advancements in NLP, accurately detecting sentiment polarity in consumer reviews remains challenging. This study addresses precise polarity identification in Amazon customer reviews, proposing an integrated methodology using advanced NLP techniques. Our approach involves initial sentiment labeling using VADER, TextBlob, and Flair, followed by feature extraction through TF-IDF vectorization, and sentiment classification using machine learning models. The Support Vector Classifier achieved the highest accuracy at 99.65%, demonstrating the effectiveness of our integrated approach. This research highlights the potential of combining NLP methodologies with traditional machine learning techniques to enhance sentiment analysis accuracy, offering valuable insights for both commercial entities and academic researchers in understanding consumer sentiments on e-commerce platforms. Future studies could explore the generalizability of this approach to other domains and languages.