Data Therapy
We need a Data Mart!
We need a Data Warehouse!
We need a Data Lake!
We need a Data Scientist!
We need DataOps!
Chances are you already have DataOps running to some degree or another. DataOps may be called different things in different organizations, but basically they are the ones that make sure the data gets to where it needs to in a timely manner.
A Data Warehouse, Mart, Lake are all projects that require a team (and time) to implement. Not to mention infrastructure, if done on-premise, or Cloud resources, which contrary to popular belief may be easy, but certainly is not cheap.
A Data Scientist can provide value and insight into your data so long as they are looking at the right data, and have the ability to enrich your data with other sources. Not to mention the resources they may need depending on your volume.
Traditional routes are tried and true, and there are lots of people within your organization that know the way and the tool-sets these traditional methods provide.
Hadoop, Spark, Cassandra, Neo4J are all "new" (to the enterprise maybe, but not to the community at large), which one solves your immediate problems? Which one is viable for the future? Should we just do traditional things, or should we start out on the road less traveled by others within our organization?
Which path do you take?
Chances are there are many more choices than just two.
Having a "sounding board" from outside your organization to discuss options in a non sales atmosphere is something that could be useful in these situations. Attend demo days of these new technologies, attend conferences, send some of your people to training.
Go, Learn things!
If you don't have the ability to do so, at least spend time speaking to someone who has done these things.
Sometimes, just having a conversation with a third party about your data may provide enough insight into where you need to go, that the path forward becomes obvious once seen.
Original article here