Data science tools and other announcements from Ignite

Program Synthesis

Microsoft's innovative approach to data wrangling through program synthesis is transforming how developers handle big data. Joseph Sirosh explains that program synthesis by example allows users to automate the tedious process of data transformation, significantly reducing the time spent on manual coding. This technology enables users to provide input and output examples, and the system generates the necessary transformation program automatically, which can then be applied at scale 1.

You can run it on a spark system at big data scale, or you can run it anywhere you choose.

--- Joseph Sirosh

This advancement is particularly beneficial in fields like finance and log analytics, where it can save up to 90% of the time typically required for data preparation 1.

   

Workbench Features

The new features in Azure Machine Learning's Workbench are set to revolutionize data handling and model deployment. Kyle Polich highlights the integration of AI-powered data wrangling directly into the Workbench client, allowing seamless data transformation 2. Additionally, the new experimentation service on the cloud facilitates the management of big data experiments, leveraging Spark and GPUs for enhanced scalability and version control.

We launched three major new features.

--- Joseph Sirosh

Furthermore, the model management feature supports deploying models to Docker containers, enabling comprehensive monitoring and updates across various environments, including cloud and on-premises setups 2.