Understand EXL XTRAKTO.AI business problem and provide data-driven solutions for those problem.Worked in Building Named Entity Recognition Model using Spacy and Bert. Worked on AWS Services i.e Lambda, Comprehend, Textract, S3(StorageBucket), Sagemaker, CloudWatch, EC2, Quicksight, SNS, SQS,Step Function,Microsoft SQL Server,ECR,ECS,Docker.Involved in Integration Module which includes Data Collection , Data Cleaning , Model Building , Creating Table in AWS RDS using Python, created stored procedure.Finetune Layout LM(v1,v2) for MRC Document Processing, extracting key information from Document by understanding document image layout structure.Deployed Machine Learning Model in Sagemaker endpoint using pre built image Container in Docker and by registering own container in ECR(BYOC). Hands-on experience in Tensor ow and Pytorch deep learning Framework for building Neural Network.Implemented Detect Drift(i.e data quality monitoring, model quality monitoring, and drift evaluation ) in production model , to detect decrease in Performance of Model and Trigger the consective action using Cloud Watch

• Understand business problems and provide data-driven solutions for those problems.• Experience in machine learning, supervised and unsupervised: Forecasting, Classification, Data/Text Mining, NLP, Decision Trees, Adaptive Decision Algorithms, Random Forest, Search Algorithms, Neural Networks, Deep Learning Algorithms• Involved in entire machine learning pipeline which includes requirement and data collection, data cleaning, data pre-processing, visualization, analysis, model training, calculating performance metrics and testing.• Help business enhance value of existing/ new products by delivering predictive models• Developed seq2seq Model using LSTM and optimisation technique which helped to improve overall performance of model by 11%• Involved in developing seq2seq Model using BERT in NLP and improve loss in model.• Built domain specific features for condition monitoring using existing edge platform.