Responsible for designing and implementing a distributed data processing architecture using Apache Spark and Python, enabling parallel processing of terabyte-scale datasets, which fed into TensorFlow and PyTorch models for real-time predictive analytics.Implemented techniques such as transfer learning and fine-tuning within TensorFlow and PyTorch frameworks, using pre-trained models on AWS SageMaker to accelerate model development and deployment across multiple use cases, along with that integrated custom machine learning models with Apache Spark to perform real-time streaming analytics, processing incoming data to dynamically adjust AI/ML algorithms deployed via AWS SageMaker.Automated data preprocessing workflows using Python scripts and Apache Spark transformations, optimizing data quality before feeding into AWS SageMaker for model training, resulting in a 25% increase in model accuracy.Developed a custom model evaluation framework in Python that interfaced with TensorFlow, PyTorch, and AWS SageMaker, automating the comparison of multiple AI/ML algorithms based on precision, recall, and F1-score metrics.Tech Stacks:: Python, Pandas, NumbPY, AWS Sagemaker, PyTorch, AutoML, Cognitive Service, TensorFlow,Spark, Kafka,

Designed and implemented machine learning models using TensorFlow and Python, leveraging Pandas for data manipulation and NumPy for numerical computations, to predict and mitigate financial risks based on real-time market data.Engineered end-to-end data pipelines on AWS SageMaker, integrating Spark for distributed data processing and Python for data cleaning and feature engineering, enabling scalable and efficient training of risk prediction models on large datasets.Developed automated model selection and hyperparameter tuning workflows using AutoML in SageMaker, optimizing TensorFlow models to improve prediction accuracy and reduce model training time for complex risk management scenarios.Utilized Spark and Python to process and aggregate massive datasets, feeding the data into TensorFlow models for real-time risk analysis, ensuring that the system could handle high-volume, low-latency data streams with minimal performance degradation.Implemented advanced statistical analysis and data preprocessing techniques using Pandas and NumPy, preparing time-series data and other financial metrics for training and evaluation of TensorFlow models within AWS SageMaker.Automated the deployment and monitoring of machine learning models using SageMaker endpoints, creating a continuous integration and deployment pipeline that leveraged AutoML to dynamically update models in response to new risk factors identified by the system.

Engineered and optimized user interfaces using HTML, CSS, and JavaScript, harnessing core Python skills tocreate seamless and visually striking web applications for HP's platforms. Ensured superior user experiencesand compatibility across devices, contributing to engaging digital services.Developed a well-structured Angular component architecture, implementing best practices such as componentmodularization, state management using Angular services, and the utilization of Angular features like reactiveforms and observables. This optimization enhanced code maintainability, scalability, and facilitated a moreintuitive user interface.Utilizing Azure Machine Learning Services and Python to create predictive models that enhanced businessinsights, improved decision-making, and optimized operational processes.Implemented Azure Blob Storage for efficient Data Storage and Management, establishing a scalable andorganized data repository that facilitated seamless integration with Python scripts for data manipulation andanalysis.Designed and optimized Angular-based user interfaces for data visualization, providing intuitive dashboards andreports that allowed stakeholders to interact with and derive insights from the results of data analytics andmachine learning models.