The advancement in Natural Language Processing (NLP), particularly in text summarization, alongside the growing volume of chat data on platforms like WhatsApp, necessitates tools to process this data efficiently. This paper presents a detailed study on analyzing and summarizing WhatsApp group chats using NLP techniques. It explores message classification, chat clustering, and summarization to extract significant information from conversations. Four machine learning models (Logistic Regression, Naive Bayes, SVM, and Random Forest) were compared for message classification, with Random Forest emerging as the best model after optimization. The classification process included manual message categorization and essential preprocessing. For chat clustering, a combined approach based on message category, author, and timestamp was developed, yielding better results than individual methods. Both extractive and abstractive summarization methods were investigated, with BERT and Gensim used for extractive summarization, albeit with limitations in capturing context. Abstractive summarization using BART and PEGASUS models from Hugging Face showed superior results, especially with the SAMSum-trained model. The findings underscore the effectiveness of the combined clustering approach and the importance of advanced abstractive summarization models for capturing context in WhatsApp group chats.