About

Mohammad Amaan Sayeed is a highly skilled Machine Learning Graduate Student at MBZUAI with a passion for all types of Machine Learning. Currently working as a Machine Learning Engineer at Chubb, he has gained hands-on experience in analyzing claim documents to identify emerging risks in organizational claims on Azure Databricks. He has parallelized ETL operations on claim documents on PySpark, Databricks, and fine-tuned RoBERTa and Longformer models to extract the cause of loss from claims using HuggingFace, Haystack. Mohammad has developed an evaluation procedure to determine the relevancy of extracted cause of loss with the intended outcome of emerging risks, using metrics such as Rouge and Bert-score. He has also fine-tuned longformer encoder-decoder based summarization models to summarize the claim documents into cause of loss summaries. Additionally, he has fine-tuned SparkNLP based models to extract the cause of loss and summarize claim documents. In the past, Mohammad has worked as an ML Intern at Chubb, where he developed a Machine Learning engine to detect and extract employee information from resume documents with a performance of ~80% on various metrics. He has also worked as a Deep Learning Researcher at Unreal AI, where he contributed to Binary Weighted Networks based framework from scratch and worked on SOTA face recognition research, outperforming several current SOTA systems. Mohammad has a Master of Science - MS in Artificial Intelligence from MBZUAI (Mohamed Bin Zayed University of Artificial Intelligence) and a Bachelor of Technology - BTech in Information Technology from Kakatiya Institute of Technology & Science. He has also completed Nanodegrees in Natural Language Processing and Deep Learning from Udacity. Mohammad's tech stack includes ML, Databricks, BERT, Deep Learning, ETL, Azure, Research Scientist, Computer Vision, TensorFlow, and PyTorch. With 1.77 years of relevant experience, he is a highly motivated individual with a strong desire to learn and grow in the field of Machine Learning.