This Recommendation system is built to recommend the book to customers based on their history of purchases. The Recommendation system comes under the Machine learning Process, which is divided into supervised and unsupervised learning, where this recommendation comes under the unsupervised machine learning. In unsupervised machine learning the recommendation engine is built using a Popularity based filtering algorithm along with a collaborative filtering algorithm. In the collaborative filtering algorithm the model-based approach is used where Cosine similarity is used as a metric machine learning and the products are recommended for the book.csv, user.csv, and ratings.csv datasets. The ML model is executed in the Jypyter notebook and the output is verified. Then a webpage for the user’s interface is designed using HTML the user interface collects the inputs from the user's dataset and transfers it to the machine model for analysis and then returns its output to the user interface page itself. Thus, the Recommendation Engine built will recommend the book to each user based on their previous purchase records and thus improve the company sale. Thus, this project focuses on increasing the sale of the company by recommending the book which the customer may like to buy. This is how this Recommendation Engine works.