Tools and infrastructure

The MLOps landscape has been developing rapidly over the last two years; many tools and frameworks have evolved as part of the infrastructural offering. You can visit https://landscape.lfai.foundation/ to see how many mainstream options have been developed to orchestrate ML, deep learning, reinforcement learning, development environments, data pipelines, model management, explainable AI, security, and distributed computing.

There is a surge in services provided by popular cloud service providers such as Microsoft, AWS, and Google, which are complemented by data processing tools such as Airflow, Databricks, and Data Lake. These are crafted to enable ML and deep learning, for which there are great frameworks available such as scikit-learn, Spark MLlib, PyTorch, TensorFlow, MXNet, and CNTK, among others. Tools and frameworks are many, but procuring the right tools is a matter of choice and the context of your ML solution and operations setup. Having the right...