Hello Alexandre, 2013/2/20 Alexandre Rafalovitch <arafa...@gmail.com>
> How does this compare to Pentaho Kettle? http://kettle.pentaho.com/ > > I am not terribly familiar with this space, so - if the question is stupid, > feel free to address it appropriately. :-) Don't worry, it's not stupid. I will make my best to try to answer to you. Pentaho Kettle is more to a product, what we have something that is more to a framework. Besides that, it seems more like an ETL or EAI tool, which you could use to extract data from some source, apply visual transformations without needing to code and load them on a database or nosql tool. It's not entity oriented and it doesn't have it's own repository. Our project address more specific cases, as we don't just extract data, we identify data on the process. Suppose for example you have several legacy systems in your company, like a CRM, an ERP and a billing system. In each of these systems, you have data related to users, that you want to process and analyse for each user. What our system does is aggregate all data for each user in a single logical place, so you can perform actions like: - suppose you have a replicated field in both CRM and ERP systems. You can choose which one to use; - as more billing data is generated, you want to aggregate the new invoices to the index, without needing to process the entire set of data again; - you might have the same data represented differently in you CRM and in your ERP, so you want to normalize it to a canonical format before choosing which one to index, or even before indexing both. This is just a possible scenario and it's not what we do as a company, ok? Of course you could do these things using a tool like Pentaho, but you would have to do them manually and you would be narrowed to their interface. Our system gives you these funcionalities as pre-built and you keep being able to use any framework or technology you choose to do other tasks that might be interesting to you in the middle. For instance, you could easilly integrate a java machine learning framework, call facebook graph api, use spring or hibernate, or anything like that in your processes. Hope I was able to answer your question, but I will be glad to try again if you have more doubts. Best regards, Marcelo
