Great tool to build both Batch and Stream BigData pipelines
What do you like best?
This tool based on well known Apache Beam framework and can be locally or on internal GCP runner in GCP cloud. Has a great set of instruments to work with Stream data and Batch data in very fast manner.
It is very easy to run pipeline and track progress of the job. Much more powerful and easy than Dataproc with native Spark.
What do you dislike?
Not so clear documentation. You should practice a lot with that to pick up configuration for your particular case.
Recommendations to others considering the product:
You should definitely consider this tool in cases of BigData processing. You will have single tool for both Stream and Batch data pipelines
What problems are you solving with the product? What benefits have you realized?
Processing huge number of JSON data both as Stream and Batch data pipelines