In the current trend, social media is one of the major platform for any kind of activities and also it became like a biggest data source in India. To track variety of issues in different domains like financial markets, education medical and other sources, nowadays variety form of social media sources are available in that twitter is one of the largest source media which have large impact in news data, financial markets, politics, economics, entertainment, health domains, etc. Pandemic COVID-19 has huge impact, particularly in the field of medical domain to generate huge data about medical issues including depression and other mental health problems also. Due to wide range of different categories of social media users, lot of misinformation, unauthorised data being generated, which causes to anxiety, hypertension, heart issues and other mental health problems like depression, brain issues, etc. In this work, we aim to develop a model using machine learning techniques to classify depression of the people. For this work, we considered tweets as an input data source performing several preprocessing techniques and developing a model to classify a person is depressive or not. Identifying predicting of the depression helps to the medical practitioners to avoid further consequences or sudden kind of shocks which causes to the deaths. In our work, the major contribution is to analyse the suitable textual tweets, pre-process and present results evidences which can help making suitable policy decisions for sustainable solutions against depression. With mental health concerns on the rise, this paper embarks on a compelling quest to unlock new methodology for the early detection and understanding of depression, ultimately fostering a brighter and healthier future for individuals worldwide.