Blossom provides a middleware to run any data flow task on different platforms.

I could execute my spark job on Flink by changing only one line of code. I also liked a lot the optimizer that can select the platform based on a cost model.

I would like to be provided with better documentation

Simple, easy and pretty rock solid. Unique in the market

It's an open source product, and pretty solid and stable. The simplicity to integrate - just add 3! lines into your _already_ existing code, and Blossom does the rest.

It is developer stage. Power Users will love the stack, but more occasional users need a few hours to understand the concept. And the documentation is pure developer focused, which needs to be improved.

What I like the most about this platform is its ease of use. One has to only express the business logic within its API, and then the platform optimizes for the underlying system usage. This way one does not need to implement system-specific details.

The initial configuration and deployment with different software versions of underlying system can be challenging. However the issues got resolved after communication. Furthermore, it would be beneficial to add support (connectors) for more systems.

Blossom provides a unique software that can automatically choose in which platform a task has to be executed to achieve best performance. It is also really cool to write your application code once and by changing only one line of code have it executed in a different platform (eg., Spark, Flink or standalone Java) regardless of where the data is stored.

Better documentation would ease deployment.

Blossom allows you to focus on your application as it handles where a task will be executed (Spark, Postgres or simple Java)

My tasks can be transferred between platforms with just minor changes

Support for SQL would be really helpful

Automated determination and training of Source data. Support for Multi Cloud with a single tool. Low code and easy to integrate.

Nothing as of now, we are still evaluating for a product.

Its ease of use and its robustness to pick the right data processing platforms.

I believe that it is missing a SQL interface, which would add to the dashboard to run analytics simpler

It is surprising how easy is to configure Databloom and how easy I could setup my federated learning models

For a few complex scenarios, the set up may require advanced acknowledge of federated learning concepts

Easy to use API, makes the training of ML pipelines faster as it can run on distributed platforms such as Spark for example.

Missing support of SQL queries, it would be useful to include this

Support for wide variety of data analytics platform was very useful.

There is a a bit of learning curve involved .. the amount of functionality is a bit overwhelming. It would be helpful to provide some more on-ramp for each use case.

Es muss nur eine Zeile Java Code eingefügt werden, keine aufwändiges Anpassen von Queries für die unteschiedlichen Endsysteme

Im Moment noch sehr tech lastig, die UI braucht noch etwas

Pues se ha utilizado para realizar pruebas en el análisis de datos en diferentes sistemas de almacenamiento (polystore).

La facilidad de uso, soporte de muchos sistemas de procesamiento de datos y su funcionalidad de Federated Learning.

Aunque la interface básica es sencilla de utilizar se puede mejorar para poder hacer los dataflows, y otro punto es que se puede mejorar la documentación.

Facilidad para consultar fuentes de datos distribuidas, interfaz grafica intuitiva, eficacia en la consulta de datos, uso amigable del sistema.

La documentación debería de mejorar, sobre todo en como soportar los casos de uso no tan comunes

Supports wide array of data processing platforms. Seamless data analytics across sources. Easy to integrate existing into existing applications

Limited support for stream data analytics and graph data processing platforms.