Sarcasm is a delicate form of expressing any opinions in a facetious way. The advent of communication using social networks has mass-produced the new avenues of socialization. It can be further said that humor, irony, sarcasm, and wit are the four chariots of being socially funny in the modern day. Among these, sarcasm is a sophisticated way of wrapping any immanent truth, message, or even mockery within a hilarious manner. In the present work, we manually extract the features of a benchmark pop culture sarcasm corpus containing sarcastic dialogues and monologues, to generate padding sequences from the matrices formed of the vector representations. We further propose an amalgamation of four Parallel deep Long Short Term Networks (pLSTM), each with a distinctive activation classifier. These modules are primarily aimed for successfully detecting the sarcasm from the text corpus through the training phase. Not limited to only that, besides generic validation testing, we also mimic the human-alike statement successions initiating from random user input seed words to produce auto-generated sarcastic dialogues, which have semantic significance in an understandable sense. Our proposed model for detecting sarcasm peaks with a training accuracy of 98.95% when trained with the discussed dataset. Consecutively it achieves 98.31% accuracy among the test cases on open source humor English literature. Our approach transcends several previous state-of-the-art works and results in sophisticated sarcastic statement generation. We also culture the probable prospects for producing even better refined automated sarcasm generation.