Gretel is proud to announce a partnership with Databricks and its designation as an independent software vendor (ISV) technology partner, meeting a core set of functionality and interoperability requirements for integrating with Databricks. Through the release of a new workflow connector, Databricks users can easily access Gretel’s high-quality, safe synthetic data as they build and customize their machine learning models in Databricks. 

‍Gretel’s synthetic data for Enterprise AI in action‍

“GenAI is fundamentally about data. And unlike GPUs, the humans that create this data aren’t getting twice as fast every year. With synthetic data, however, Gretel helped us overcome this challenge. We were able to design and iterate on large, high-quality, targeted datasets that have already dramatically improved our models at essential tasks like text-to-sql. We’re very excited to continue honing our models in partnership with Gretel. Thanks to Gretel, our dataset creating can now keep up with our scientists and compute.”  - Jonathan Frankle, Chief Scientist, Neural Networks, Databricks

Gretel and Databricks make it easier than ever to build custom models for Enterprise AI‍

The recent advancements in generative AI have democratized machine learning, bringing the power of large language models to every business. In order for enterprises to utilize these tools for improving their own core business offerings, they must safely train models on their own private data. Access to this high-quality, safe training data, has traditionally been a bottleneck, but synthetic data solves this by generating on-demand data without the privacy risks or data access limitations. 

Together, Gretel and Databricks offer an easy way to create custom, proprietary models based on a business’ own unique needs. With Gretel, you can generate safe, synthetic datasets to train the latest state-of-the-art open-source models—all within the Databricks platform through the Mosaic AI Foundation Model APIs. While customizing the latest model can seem daunting, the first step is curating the right dataset that can be used to evaluate different methods like RAG, prompt-tuning, and fine-tuning. Learn more about our partnership.‍

Seamlessly integrate synthetic data into your Databricks workflow with Gretel’s new connector

Gretel customers can now access data stored in Databricks’ Data Intelligence Platform with the launch of Gretel’s new Databricks connector. The connector is the latest addition to Gretel’s third-party data connectors, and makes data stored in Databricks easily accessible on the Gretel platform with just a few clicks. With this latest connector release and Databricks notebook, companies can generate high-quality synthetic versions of their data and automatically add it to their Databricks data pipelines—expediting time-sensitive data and AI projects. 

Get started with Gretel and Databricks today

Start generating synthetic data in the Gretel Console and synthesize data stored in your Databricks Data Intelligence platform in just a few clicks with Gretel’s connector.

Headed to Databricks’ Data + AI Summit? Meet with us!