Roadmap

Here are some things we are working on/have on the roadmap

• native support for Snowflake using Snowpark. This will let Snowflake users run Zingg without an external Spark cluster
• stop words detection and configuration per column. this will help improve the matching accuracy (0.3.4)
• abbreviation detection and matching
• cli for ease of use
• automated unit and e2e testing (ongoing, first release 0.3.4)
• exposing model statistics and convergence so that people learn when to stop
• suggesting or figuring out which fields should be used in matching
• more transformations and experiments to build better models
• geo matching
• unsupervised models
• data stewardship of the results
• easier blocking and matching functions through Java/scala/Python