Databricks Expands Presence in AI and LLMs with Lilac Acquisition

And three. This week, Databricks announced the acquisition of Lilac, a startup that entered the artificial intelligence market last year with the goal of developing an open source tool that allows data scientists to see and to quantify their data sets. This is the third major acquisition in this area for the firm, after the acquisition of Einblick and Mosaic ML. In this case, the amount of the operation has not been revealed.

Lilac can be used for a range of use cases from evaluating the output of large language models (LLMs) to understanding and preparing unstructured datasets for model training. With the integration of the Lilac tool into Databricks, the American firm intends to help companies accelerate the development of generative AI applications using their own data.

A scalable and easy-to-use solution

Founded by Daniel Smilkov and Nikhil Thorat, formerly of Google, Lilac has stood out in the field of generative AI by offering a scalable solution that facilitates interaction with data. With an intuitive user interface and AI-powered features, Lilac offers a wide choice of functions. This includes mining data packets, creating data categories using human sentiments and classifiers, and adapting datasets based on previously defined classifications.

The team behind Lilac specifically designed their product to enable the analysis of model results for bias or toxicity, as well as the preparation of data for RAG and fine-tuning or LLM training.