
- CEO and Co-Founder, Databricks15
- Adjunct Professor, University of California, Berkeley1
Ali Ghodsi is the co-founder and CEO of Databricks, a role Ghodsi took in January 2016 after serving as the company's VP of Engineering and Product Management. Databricks' speaker biography describes Ghodsi as one of the creators of the open-source Apache Spark project and as an adjunct professor at UC Berkeley. Ghodsi co-authored the 2011 NSDI paper on Mesos, a platform for fine-grained resource sharing in the data center, written by a UC Berkeley team. Ghodsi is also among the bylined authors of Databricks' 2024 posts announcing the DBRX language model and the open-sourcing of Unity Catalog.1234
At a glance
- Current roles
- CEO and Co-Founder +1 more
- Companies
- Databricks, UC Berkeley
- Sector
- Company
- Fields
- Datasets and data, Language models
- Education
- UC Berkeley
- Work listed
- 4
- Sources cited
- 9
- Profile reviewed
Career
Ghodsi is a co-founder of Databricks. Before becoming chief executive in January 2016, Ghodsi served as the company's VP of Engineering and Product Management. Databricks' speaker biography also lists Ghodsi as an adjunct professor at UC Berkeley and a board member of the university's RiseLab, and notes a PhD in distributed computing from KTH Royal Institute of Technology in 2006.1
In July 2023 Databricks announced it had completed its acquisition of MosaicML; Ghodsi, quoted as co-founder and CEO, described it as a way for customers to train and deploy their own language models on their data. Ghodsi is among the bylined authors of the March 2024 post introducing DBRX, an open mixture-of-experts language model, and of the June 2024 post announcing the open-sourcing of Unity Catalog.5346
Research
Databricks' biography notes that ideas from Ghodsi's academic research on resource management, scheduling and data caching have been applied in Apache Mesos and Apache Hadoop. At NSDI 2011 Ghodsi co-authored the Mesos paper on fine-grained resource sharing and was first author of "Dominant Resource Fairness", which addresses the fair allocation of multiple resource types in a shared cluster.127
Ghodsi is described as one of the creators of the open-source Apache Spark project, which is developed under the Apache-2.0 license. With Michael Armbrust, Reynold Xin and Matei Zaharia, Ghodsi co-wrote the CIDR 2021 paper "Lakehouse", which argues for data platforms built on open direct-access file formats with built-in support for machine learning and data science.189
Work and contributions
Apache Spark Open-source code
Databricks' speaker biography names Ghodsi as one of the creators of the open-source Apache Spark project. The apache/spark repository is published under the Apache-2.0 license.18
Unity Catalog Open-source code
Co-author, with Matei Zaharia, Reynold Xin, Arsalan Tavakoli-Shiraji and Patrick Wendell, of the June 2024 Databricks post announcing that Unity Catalog was being open-sourced under the Apache 2.0 license and hosted at LF AI & Data. The unitycatalog/unitycatalog repository is Apache-2.0.46
Mesos Published research
Co-author of "Mesos: A Platform for Fine-Grained Resource Sharing in the Data Center" (NSDI 2011), whose eight authors are all listed at the University of California, Berkeley.2
DBRX
One of six bylined authors of Databricks' March 2024 post announcing DBRX, a mixture-of-experts language model built on the MegaBlocks research and open-source project.3
Timeline
- Co-authored the Mesos paper on fine-grained resource sharing in the data center at NSDI 2011.2
- Presented "Dominant Resource Fairness" as first author at NSDI 2011.7
- Became chief executive of Databricks after serving as VP of Engineering and Product Management.1
- Co-authored the CIDR 2021 paper describing the Lakehouse architecture for open data platforms.9
- Databricks announced the completion of its acquisition of MosaicML, with Ghodsi quoted as co-founder and CEO.5
- Co-authored the Databricks post announcing the open DBRX language model.3
- Co-authored the Databricks post announcing the open-sourcing of Unity Catalog.4
Where they work
- Databricks 11 people in the wiki
- UC Berkeley 27 people in the wiki
Related people
- Jonathan FrankleAlso at Databricks
- Andy KonwinskiAlso at Databricks
- Nikhil ThoratAlso at Databricks
- Arsalan Tavakoli-ShirajiAlso at Databricks
- Patrick WendellAlso at Databricks
- Pradeep DasigiLanguage models · Datasets and data
- Kezhi KongLanguage models · Datasets and data
- Rangan MajumderLanguage models · Datasets and data
- Vaishaal ShankarDatasets and data · Language models
- Guilherme PenedoDatasets and data · Language models
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
9 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026
- Ali Ghodsi — Data + AI Summit 2026 speaker page
- Mesos: A Platform for Fine-Grained Resource Sharing in the Data Center (NSDI '11)
- Announcing DBRX: A new standard for efficient open source LLMs
- Open sourcing Unity Catalog (Databricks blog)
- Databricks Completes Acquisition of MosaicML (press release)
- unitycatalog/unitycatalog repository (Apache-2.0 license)
- Dominant Resource Fairness: Fair Allocation of Multiple Resource Types (NSDI '11)
- apache/spark repository (Apache-2.0 license)
- Lakehouse: A New Generation of Open Platforms that Unify Data Warehousing and Advanced Analytics (CIDR 2021)