91制片厂

What differentiates Databricks Data Intelligence Platform from other similar products?

It's broad end features

What is your favorite aspect of this product?

Easy to manage, support of multiple language and can be interacted with different Azure services

What do you dislike most about this product?

Nothing specific

What recommendations would you give to someone considering this product?

Go for it

What differentiates Databricks Data Intelligence Platform from other similar products?

Great expertise that mixes ml-Runtimes 鈥� MLFlow and Spark. The flexibility to use Python, and SQL seamlessly in one platform. Since databricks notebooks are saved as python scripts in the background it鈥檚 superb to have both notebook and script experience and synchronize to git. Regular jobs, pre-installed ML environments, an abundance of documentation, and example use cases. There are tons of best practices obtainable through the Spark AI conference also.

What is your favorite aspect of this product?

I actually like that we can have queries. We鈥檝e solely began to review them properly for each account, however there are limitless opportunities once we begin digging. It provides multi-cloud support across AWS etc. New options are sharply released every quarter. The UI is comparatively easy compared to AWS EMR or different similar products. The approach that works with the delta format. Providing tons of potentialities without the requirement to own a dedicated database administrator. It鈥� conjointly nice to talk concerning their flexibility.

What do you dislike most about this product?

Debugging code and using interactive applications outside databricks approved features is tough. It鈥檚 difficult to get a grasp of the documentation for beginners to the platform. Their Spark install is not forever compatible with third party tools (ex: geospatial) causing a delay in some package availability till an extra release. Some gotachs with distributed model coaching (Horovod). I鈥檇 prefer to access additional access to making queries and investigating many things. Generally, I get glitches with the account, it freezes and I can鈥檛 see the data.

What recommendations would you give to someone considering this product?

It鈥檚 highly efficient in executing queries, analyzing data, and performing complicated table joins if workloads are properly set up across high-performing clusters. The simplest part is looking for information on a couple of customers鈥� accounts that we cannot see on the other system we use. Therefore I like to recommend this.