This research is concerned with domain-specific named entity recognition in which certain keywords are withdrawn from the phrase and sentences of the words or paragraph. The project in hand has been effective in different sectors like the insurance companies and hospitals for withdrawing the required keywords in a small amount of time. For instance, a medicine that belongs to a specific keyword is extracted easily and the same occurs with necessary keywords in the case of insurance companies. Advancements in deep neural networks have been established to build the named entity recognition model (NER Model). We have used Spacy and natural language processing for keyword extraction. The future scope of the project is to use LSTM for domain-specific keyword extraction and transfer learning using NLP. A large dataset is required to build the system for keyword extraction since dataset have always been the challenging part of such a system.