Developed machine learning models in order to classify/predict information requests and sentiment. Worked cross-functionally to develop LDA topic modelling and similarity between documents with increased efficiency. Implemented face recognition, cosine similarity between faces and a mapping feature. Learnt and improved the performance of existing Sequence to Sequence model by 7 % for text recognition in terms of character and word level. Built the CNN architecture for Multi-label document classification using deep learning techniques, thus improving the overall performance by 5%. Refactored the code from jupyter notebooks to python modules. Developed SQL scripts(with statistical analysis), extraction, integration, optimizing analysis time and normalization of data with ETL process in Talend workspace. Thus, providing the desired insights to our clients. Development of batch scripts in python for operating over millions of records and data migration ensuring performance and efficiency. Automation and maintenance of Talend jobs for ensuring end to end workflow. Algorithms used: Logistic regression, Multinomial Naive Bayes, SVM, ensemble/stacking methods, CNN, Facenet, LSTM/BiLSTM, GRU/BiGRU.