Release Notes for RapidCanvas Oct 25, 2023

New features

This release includes the following features:

AI Guide on Project Canvas

  • Introducing an AI guide that offers users insights on machine learning project implementation based on their specific use case and industry.

  • Engage in a dynamic conversation by asking follow-up questions.

Rapid Model Recipe

  • Presenting a simplified way to generate models. Upload a flat file and let the Rapid Model Recipe take care of the rest.

  • Current support includes Binary Classification, Regression, and Multi-Class Classification.

  • Stay tuned for upcoming support for Time Series Forecasting, Clustering, and Anomaly Detection in our next release.

Enhancements

Data Import

  • Seamless fetching of table-form data from MongoDB and Snowflake sources. Import this data directly onto the canvas for transformative operations.

  • Integration with Fivetran connectors now permits data imports directly into projects.

Other Enhancements

  • Interface adjustments on RapidCanvas for Windows OS users

  • Enhanced data consistency and accuracy when handling files with special characters or separators.

  • Introduced job count linked to a project under the Environments tab for more transparency.

Bug fixes

These are the issues resolved as part of this release:

UI and Design

  • Long filenames no longer overflow in the “Create New Dataset” window.

  • Various alignment and UI improvements across Recipe, Tenant, and Environment pages for a streamlined experience.

Data Handling

  • Resolved encoding/separation errors for uploaded files, offering on-the-spot corrections.

  • Addressed inconsistencies with data type and ontology displays.

Functionality

  • Improved feedback mechanisms for data uploads, ensuring users receive accurate success/failure messages.

  • Better validation across the environment page for user inputs.

  • User alerts have been optimized for various platform actions, such as environment relaunches.

  • Enhanced error reporting for recipes, ensuring the UI reflects backend statuses accurately.

Environment and Resources

  • Improved tooltips, feedback, and options related to environment status, launch, and usage.

  • Addressed environment naming and package editing restrictions for better platform consistency.