Please note the date of this post. Azuqua is developing fast and their features change quickly.
At my previous company I worked extensively with an integration and automation platform called Azuqua and loved it. Azuqua was relatively intuitive with an understanding of branching and looping logic, so development was fast and their support team was fantastic. The UX was also enjoyable to spend all day starting at. However, when i moved companies and was selecting an integration and automation platform to build our core systems integrations on I did not choose Azuqua. There two major flaws and two major nuisances that pushed me away from it.
Azuqua does not have environments. You can build a process connected to the test instances of your applications until you get it right, but at that point there is no simple way to push it to work with your production environment. You can clone it and update every connector call in your process to your production credentials individually, but every time you want to make an update you want to test you have to do this. You also frequently have to reselect a lot of options other than just the credentials when updating a connector to production.
We ran into stability issues with Azuqua. Azuqua pushed updates on Fridays and did not do a good job announcing what was changing. This left us in situations where Monday morning we would realize something had been running incorrectly, or not running at all over the weekend because of a change made to Azuqua that was pushed on Friday and we were not notified of in advance, so we would make sure out processes would continue to work. This is not out of the norm for a startup and will get better over time, but it is not what you want in a platform responsible for transmitting key business data.
Mapping a large number of fields took a long time. In order to map a field in Azuqua you have to drag it from the source to the place where you are going to use it. This is fine if you are retrieving a few fields, but if you are mapping 100 fields this becomes painful. If you are mapping 100 fields from the beginning of a long process to the end this is even more painful because you have to horizontally drag the property every time until you reach the end of the process.
Somewhat complex math requires an absurd number of cards. In Azuqua every operation is a card. This is no different for math. Every mathematical operation requires 1 card. This means that doing even relatively simple math becomes verbose and confusing in Azuqua because you have to be aware of the order of operations, result of every operation within the equation and what to do with it next. This is one of the few places where using Azqua is far more painful than writing the code by hand would be.
I look forward to Azuqua resolving these issues because outside of them it was a pleasure to develop on. I am currently using Dell Boomi because it solves the stability and environment issues, but it is less intuitive, not very pleasurable to look at, and the support is awful.