Databricks is a global company focusing on data and AI. At the core of Databricks is the Databricks Data Intelligence Platform which allows entire organizations to use data and AI to power a wide range of business use cases. It's built on a lakehouse to provide an open, unified foundation for all data and governance and is powered by a Data Intelligence Engine that understands the uniqueness of the organizations’ data. Databricks simplifies and accelerates enterprises' data and AI goals by unifying data, analytics and AI on one platform. Its key mission is to assist data teams in addressing some of the world's most challenging problems.
Do You Manage Peer Insights at Databricks?
Access Vendor Portal to update and manage your profile.
Strong compute for data intensive workloads. It can process large volumes of data quickly for follow up tasks like insights or agentic decision making. The suite is also quite complete (not fully). With MLFlow a quick and easy development studio can be set up, while unity catalog allows fine grained enterprise wide data sharing.
What I personally like most about it is its unified architecture that seamlessly combines data engineering, analytics and AI in one platform and it also values how it simplifies workflows, enabling faster insights and more efficient decision making across organization.
1. The sheer time savings a tool like this offers is immense, 2. The democratisation of data is really helpful for less technical folk to get information at their fingertips, 3. Being able to describe outcomes in natural text rather than SQL is very useful.
The solution is quite costly, there are cheaper alternatives in the market that work just as well for certain use cases. If your use case has little data consumption, DBX is just to expensive. Also we are struggling with combining stateful agents with DBX, as it requires external state persistence. As such, if this is not available, an extra component needs to be introduced in the solution.
It can be difficult for beginners to get started due to its complexity. It can also become expensive if usage isn't carefully managed. Optimizing performance often requires experienced users.
1. Regional availability of Azure for Genie Agent is patchy, 2. The number of tables you can give to Genie is limited, 3. Cross-region data processing doesn't guarantee which region your data will be processed within.