Tomorrow marks the start of AWS re:Invent 2019. I am excited to go and deepen my expertise with machine learning on AWS.
My general goals for re:Invent are the following:
Improve my understanding of serverless services (e.g. Lambda, S3, Glue, API Gateway, Kinesis, Athena) and how they can be used to build and monitor flexible, scalable machine learning pipelines. I’d like to build and document 1-3 sample pipelines using batch, streaming, and/or unstructured data (e.g. text, images).
Improve my understanding of managed services for ML/AI, including SageMaker and EMR, to enable rapid data science development with minimal concern for infrastructure. I’d like to train, tune, and deploy at least 2 models using Spark and/or Scikit-learn.
Understand the current state of BYOA (bring your own algorithm) and how to easily deploy inference endpoints from models built in R and/or Scikit-learn. I’d like to train and deploy 1 model built with R.
Pass the Machine Learning certification exam.