We didn't consider other options besides Lumada because we are members of an auto parts trade association, and they were using the Pentaho tool before it was Hitachi to do some ETL tasks. They recommended it, so we started using it. I evaluated a couple of other ones, but they cost more than we were willing to spend to try out this type of solution. Once we figured out what it could do for us, then it's like, "Okay. Now, we can do some real work here."

While the capabilities of Pentaho are good enough for light work, I've started using Alteryx Designer, and it is so much more robust in everything that you can do in real time. I've also used SSIS.

When you run something in Pentaho, you can click on it to see the output of each one, but it's hard to really change anything. For example, if I were to query data from a database and put it into a "select," if I wanted to reorganize within the select based on something like the first initial of someone's name, it provided that option. But when I would do it, sometimes it would throw an error and I'd have to run the feed again to see it.

The nodes, or the components, in Pentaho can probably do about 70 percent of what you can do in Alteryx. Don't get me wrong, Pentaho worked for what we needed it for, with just a few quirks. But as a data engineer, I'm always interested in and excited to work with new technologies that may offer different benefits. In this case, one of the benefits is that each node in Alteryx has many more capabilities in real time. I can look at the data that's coming into the node and the data that's going out. There was a way to do that in Pentaho, if you right-clicked and looked, but it would tell you the fields that were coming in and out and not necessarily the data. It's nice to be able to troubleshoot, on the spot, node-by-node, if you're having an issue. You can do that easily with Alteryx.

In addition to being able to look at data coming in and out of the node, you can also sort it easily and filter it within each data node in Alteryx, and that is something you can't do in Pentaho.

Another cool thing with Alteryx, although it's a very small difference, is that you don't have to save the workflow before you run it. Pentaho forces you to do that. Of course, it's always good to save.

We did evaluate SSIS since our database is based on Microsoft SQL server. SSIS comes with any purchase of an SQL Server license. However, even with SSIS, there were some limitations. For example, if you want to build a package and reuse it, SSIS doesn't provide the same kinds of abilities that Pentaho does. The amount of reusability reduces when we try to build the same thing using SSIS. Whereas, in Pentaho, we could literally reuse the same code by using some of its features.

SSIS comes with the SQL Server and is easier to maintain, given that there are far more people who would have knowledge of SSIS. However, if I want to do a PCP encryption or make an API connection, it is difficult. To create a reusable package is not that easy, which would be the con for SSIS. 

I did not evaluate other options.

I used to work with a reseller of Pentaho. That is why I started working with it. Also, I did some training for Pentaho at the company that I used to work for in Argentina, where we were a Platinum reseller. 

Pentaho is easy to use. You don't need to install anything. You can just open the script and start working on it. That is why I chose it. With Informatica, you need to do a server installation, but some companies might not allow some installation in their production or normal environment.

I feel pretty comfortable using the solution. I have tried to use other tools, but I always come back to Pentaho because it is easier. 

Pentaho is open source. While Informatica is a very good tool, it is pretty expensive. That is one of the biggest cons for the data team because you don't want to pay money for tools that just only help you to work.  

I just use the ETL tool. For data visualization, we are using Power BI. For data storage, we use SQL Server, Azure, or Google BigQuery.

We are just using the open-source application for ETL. We have never looked into other tools of Hitachi because they are paid.

I know other companies who are using Alteryx, which has a friendlier user interface, but they have fewer tools and are more difficult to utilize. My wife uses Alteryx, and I find it is not as good after I used Lumada because they have more solutions and it's open-source. Though, Alteryx has more security and better support.

I also use Talend Data Integration. For me, Lumada is straightforward and makes it simpler to have transformations as drag and drops. Comparing Talend and Lumada, I think Lumada is easier to use, more than Talend. The comprehension needed for these tools is less with Lumada with than Talend. I can learn Lumada in a day and proceed with my transformations, using some tutorials, since Lumada is easier to use. Whereas, Talend is a more complex solution with more complex transformations.

In Talend's open version, i.e., free version, you won't have a Talend server to deploy models. Thus, you deploy Talend models on the server. If you want to schedule some transformation, then you need to use the operational system where you have infrastructure to run transformations and deploy them. For example, in Talend, we deployed a data model in Talend, but we needed to use Windows Scheduler to also schedule the packets in Talend to process the data in the free version of Talend. Whereas, in the free version of Lumada, we already had it based on the web server. Therefore, we can run our transformations and deploy them on the server. We can schedule in a web interface, which guides us with scheduling the data and checking our logs to see how many transformations we have at a time. This is the biggest difference between Talend and Lumada.

I've used Tableau and other reporting tools, but Tableau sticks out because the reporting tool is much nicer. Tableau has its drawbacks with the ETL, because you can only use Tableau datasets. You have to get data into a Tableau file dataset and then the ETL part of it is stuck in Tableau forever.

If we could use the Pentaho ETL and the Tableau reporting we'd be happy campers.

I have had experience with other solutions, but for the last project we did not evaluate other options. Because we had previous experience with Pentaho Data Integration, it was pretty much a no-brainer to use it.

When we made the choice, it was between Microsoft, Hitachi, and Cognos. The deciding factor in going with Hitachi was its better support for open-source databases and data stores. Also, the functionality of the Community version was what was needed by most of our customers.

Apache has a project going on called Apache Hop. Because Pentaho was open sourced, people have taken and forged it. They are really modernizing the solution. As far as I know, Hitachi is not involved yet. I would highly advise them to get involved in that open-source project. It will be the next generation of Pentaho. If they get left behind, they're not going to have anything. It would be a very bad move to just ignore it. Hitachi should not ignore Apache Hop.

We looked at Tableau, Pentaho and an IBM solution. In the absence of Pentaho, we would have gone with either Tableau or building our own custom solution. When we were figuring out what third-party tool to use, we did an analysis and a bunch of other tools were compared. Ultimately, we went with Pentaho because it did have a wide variety of features and functionalities within its reports. Though I wasn't involved, there was a cost analysis done and Pentaho did favorably in terms of cost.

For the product that we use Pentaho for, I think we're happy with their decision. There are a few other products in our product suite. Those products ended up using Tableau. I know that there have been discussions about considering Tableau over Pentaho in the future. 

There was an evaluation performed, however I was not involved in it.

I had the chance to test SAS Data Integration but I didn’t fall in love with it like I did with PDI. I think that PDI is easier to use and you can do much more with PDI than with SAS.

Before using Pentaho, I analyzed other products to understand what is the best ETL product. I tested Talend and Oracle Data Integrator. Oracle Data Integrator is a little bit more difficult to understand, how it works.

So, I preferred Pentaho Data Integration because you just have to drag and drop the block, draw a line to connect the block, write the query, and connect to the DB. There's nothing else you need to do. For Oracle Data Integrator, and also for Talend, you spend more time installing the product. By contrast, with Pentaho, you just have to copy the folder, launch the product, and then you just need the Java machine and it works.

No.