Databricks Hones AI Strategy with $100 Million Arcion Buy


By: Mary Jander

Databricks, purveyor of the data lakehouse approach to cloud-native data management, will purchase Arcion, a startup partially funded by Databricks that specializes in replicating data across multiple data management platforms. The transaction is worth about $100 million, according to Databricks.

The acquisition is a further move by Databricks to equip its data platform with the tools necessary to enable enterprises to not only manage their data but to deploy it in artificial intelligence (AI) models and applications tailored to their specific requirements.

Back in June, Databricks spent about $1.3 billion to acquire MosaicML, a San Francisco-based firm founded in 2021 to create large language models (LLMs) adaptable to multiple use cases. That merger should allow Databricks customers, which reportedly number over 10,000, to generate LLMs using the data stored in Databricks’ lakehouse environment. The lakehouse model devised by Databricks combines data warehousing with the storage of data in its original state, regardless of its structured or unstructured format. With the addition of Arcion, customers now have a means of porting data from multiple enterprise sources into the Databricks products. Databricks CEO and co-founder Ali Ghodsi told CNBC that Arcion will be integrated into Mosaic ML.

"To build analytical dashboards, data applications, and AI models, data needs to be replicated from the systems of record like CRM, ERP, and enterprise apps to the Lakehouse," said Databricks CEO and co-founder Ali Ghodsi, in the press release. "Arcion's highly reliable and easy-to-use solution will enable our customers to make that data available almost instantly for faster and more informed decision-making.”

Traditionally, bringing enterprise-specific data into the AI sphere has been notoriously challenging and cumbersome, so if Databricks can meld its acquisitions, it will solve a couple of major roadblocks to meaningful deployment of generative AI by enterprise customers.

More on Arcion

Arcion, based in San Mateo, Calif., was founded in 2016 as BlitzzIO by Rajkumar Sen (ex-Oracle, now CTO) and Miryana Joksovic. By February 2022, the company had changed its name, acquired a new CEO and co-founder, Gary Hagmueller, formerly CEO of graph database startup Dgraph, and publicly released its data platform. Arcion also announced then that Databricks participated in Arcion’s Series A round of $13 million led by Bessemer Venture Partners.

Arcion’s platform boasts a library of over 20 data connectors capable of moving data from Oracle, SAP, IBM, and other databases into other environments, including Databricks’. The platform does this with scalable change data capture (CDC), ensuring that any changes to data are updated across environments in real time. Arcion claims it can transfer 1 Tbyte of data in under 30 minutes.

Databricks Triumphant

In September 2023, Databricks revealed its own $500 million funding round on a valuation of $43 billion, bringing the total it’s raised to about $4 billion. The company also said it has reached a $1.5 billion revenue run rate and realized over 50% year-over-year growth. Its Series I round was led by T. Rowe Price along with a fleet of influential investors, including NVIDIA.

All this news has prompted speculation that Databricks may be close to going public sometime soon. But CEO Ghodsi deflects the question publicly, stating only that Databricks is interested in solid growth based on meeting customer requirements through product development and acquisitions.

So far, the strategy is clicking into place like Legos.

Futuriom Take: In acquiring Arcion, Databricks has put in place another key element of enterprise generative AI, namely a way to bring enterprise-specific data into large language models to meet the needs of specific industries.