Learner Profiles

Alex from Environmental Science

Alex is a PhD student studying climate patterns and wants to automate the process of training predictive models for temperature and precipitation forecasting. They have experience working with Python and Jupyter notebooks but have never used AWS before. Alex wants to learn how to store and access large climate datasets in S3 and train machine learning models using SageMaker. They are particularly interested in quickly testing multiple models to compare predictive performance.

Jamie from Bioinformatics

Jamie is a bioinformatics researcher analyzing genomic data for disease prediction. Their team has been using local computing resources but is facing limitations with processing large datasets. They are interested in learning how to use SageMaker to efficiently train deep learning models, leverage built-in hyperparameter tuning, and run multiple model architectures in parallel to find the best-performing approach.

Riley from Social Sciences

Riley is a sociology researcher working on predictive models to analyze trends in public opinion surveys. They have primarily used R and SPSS but are interested in transitioning to Python and cloud-based ML workflows to scale up their modeling experiments. Riley wants to understand how SageMaker notebooks can be used as controllers for model training and how to leverage SageMaker’s automated hyperparameter tuning to accelerate their research.

Taylor from Healthcare Informatics

Taylor is a medical researcher using ML to predict patient outcomes based on electronic health records. They have worked with tabular data in pandas and SQL but have never used AWS. Taylor wants to understand how to securely manage healthcare data in S3 and run predictive models in SageMaker while ensuring compliance with data regulations. They are also interested in experimenting with Hugging Face models via SageMaker for text-based medical applications, such as processing clinical notes.

Jordan from NLP and Large-Scale AI

Jordan is a researcher working on large-scale natural language processing (NLP) models, including fine-tuning LLMs for domain-specific applications. They are interested in running large multi-billion-parameter models using distributed training and need to efficiently leverage multiple GPUs across multiple instances. Jordan wants to learn how to use SageMaker’s distributed training capabilities, parallelized inference, and model deployment strategies to handle high-throughput AI workloads