We are a finanical firm with global offices, but a small IT team. Having applications with minimal support is key. KNIME fits that without problems.
We started with Alteryx about two years ago. We discovered KNIME through academic literature and tested it. We took a quick look at PowerBI, Pentaho and RapidMiner also. After that it was a quick switch to KNIME, and an equally quick move to the server version. We now run KNIME within a small team, with IT support solely to maintain thge virtual machine the KNIME server runs on. We are close to considering the web portal license to extend its use.
We use KNIME and connect to SQL Server, Oracle, Neo4j for databases. We connect to Bloomberg for live data feeds, to the web using Selenium and Palladium nodes, and to Tableau for visualisation. Analytics is done natively, and using python. We have used it as the backbone for text analysis, data analysis, image analysis, and web analysis. File system and system control is done via natively through the knime nodes and with java through the snippets. We are not java programmers but it is easy to pick up enough to work. Job scheduling is controlled by the KNIME server, and job monitoring is done by a KNIME flow coupled into Tableau.
We have no problems with the desktop Analytic Platform. It did take a little work to get the KNIME server installation going, especially compared with Confluence or Tableau which fully installed themselves. We are a Windows shop and the Server installation for Windows is not automated. However, Iris and the KNIME team was very helpful in guiding us, and once started it just runs. I have no doubt they will be very helpful.
We have not needed any support from IT beyond hardware and some initial database driver setup, largely to find jdbc drivers. Connection with python was straightforward once we got python properly installed. The fact it is an open source has not been a problem. Updates are straightforward, we have been through major and minor version updates. We manage them ourselves, and not by the IT team.
Our Server workflows are on network drives and are automatically covered by global backup policies, as are our individual workflows. We have a disaster recovery policy which probably needs reviewing as we have upgraded from the time it was put in place. From an IT support perspetive, our IT team never need to do anything with it.
I'm sure you have tested out KNIME's analytic abilities and ETL power. My tuppence's worth is KNIME doesn't try to be the best, it becomes the best by allowing you to integrate with something that actually is the best for the task at hand. Being open source makes this possible. It's an integration tool. The others try to be the best by themselves, and fall far behind. Open source is what makes this possible.
My vote, go for KNIME, forget the others. Your IT team will worry they don't have a job!