Databricks Valuable Features

Of the available feature set, I like the Imageflow feature a lot. It is very interesting. It gives me clarity on the execution of a process. I can draw the complete flow from start to finish in the exact way that I want it to execute. It is more visual and it is also easier for the people in businesses where I make presentations to understand. When I demonstrate a process to a business and show them the approach I am taking using code and technical language, then of course not many are going to understand that. But when I show them the process in terms of the graphical layout Imageflow helps provide, then they will be able to understand it much easier. They understand why I am choosing a particular way of executing the process and why I am taking certain steps in the way I have chosen to do it. The point is to help other people understand the solution more clearly. View full review »

Immense ease in running very large scale analytics, with a convenient and slick UI. This saved us from having to tweak, tune, dive into deeper abstractions, get involved in procurement, and also having to wait for other workloads to run. The built-in optimization recommendations halved the speed of queries and allowed us to reach decision points and deliver insights very quickly. The Delta data format proved excellent. Databricks had already done the heavy lifting and optimized the format for large scale interactive querying. They saved us a lot of time. View full review »

The most valuable feature is the ability to switch loads between multiple clusters. Automation with Databricks is very easy when using the API. The ability to write code and SQL in the same interface is useful. It is easy to connect notebooks to a cluster. There are a large number of inbuilt functions that help to make things easier. View full review »

Valuable features would have to include the Notebook for piping some models and the future of executing the notebooks in parallel, in batches, which is also something that we use. And we use the Notebook on Spark with Python. View full review »

I found that PySpark is the most useful tool. It uses in-memory calculation and when you want to run a model it does it very quickly. We used to use Python and when we migrated to PySpark the performance was much better. View full review »

The time travel feature is the solution's most valuable aspect. View full review »

The most valuable aspect of the solution is its notebook. It's quite convenient to use, both terms of the research and the development and also the final deployment, I can just declare the spark jobs by the load tables. It's quite convenient. View full review »

Databricks is based on a Spark cluster and it is fast. Performance-wise, it is great. This solution has very good machine learning libraries built-in. The support for big data is good. View full review »