1. Academic paper contributions related to Natural Language Processing (NLP) and Reinforcement Learning.2. Design and implement different complex rule-based algorithms for information retrieval tasks on large text data sets such as academic papers or financial state of advice documents.3. Experiment and apply the pretrained Google BERT model, which is a language model trained on a large number of documents with millions of parameters, to measure semantic similarity between words or sentences.4. Experiment and apply the state-of-the-art pretrained Google T5 model, which is a transfer learning model trained for multiple NLP tasks, to recommend summaries and questions for text data.5. Experiment and apply Sequence-to-Sequence (Seq2Seq) with beam search algorithm and attention mechanism, which is the deep neural network model trained on Stanford Question Answering Dataset (SQuAD), to generate questions and answers from paragraphs.6. Transform AI research to AI applications in various domains such as recommendation system, finance and education with extensible and scalable software architectures running on Amazon Web Services (AWS), CPUs and GPUs machine.7. Develop best-practice machine learning life cycle which includes processing data, training and evaluating model, and deploying the best one.