In recent decades, data science played one of the vital roles in revolutionizing the financial industry. This paper leverages market historical data, the company’s fundamental data, recent trends, feature engineering, and various pre-trained machine learning algorithms to construct and optimize high-return portfolios. This experiment emphasizes minimizing the risks so that ROI can be increased from the portfolio. The concept of information retrieval is used to get the company’s fundamental and price trend data from financial sites. After data gathering phase, in-depth exploratory data analysis is performed to draw the facts from the dataset(s). Next, comes feature engineering with the Capital Asset Pricing Model (CAPM). Further, MPT theory is applied to check the similar returns from the assets. Clustering is used on the pricing dataset to make two different portfolios; further, these portfolios are merged with the company’s fundamental data to get an overall return. Various regression models like Linear Regression, Decision Tree, SVM, Lasso Regression, Random Forest, Ridge Regression, Elastic Net, SGD, and Artificial Neural Networks are used on the merged dataset to study the comparative performance of the portfolios.