What is our primary use case?
All of these things totally depend upon your business logic. If you want to remove duplicates or you want to implement some kind of format for postal addresses, like keeping pin codes in this format, you can customize your code accordingly and then you can get consistency in the data.
Since the way everyone is moving towards AI, my suggestion will be to lean more towards natural language processing. So rather than writing proper SQL, the way Snowflake is processing or Cloudera is processing things, they should also have a chatbot or something similar where we can simply write things, it understands that particular thing and enhances that particular thing at the back end.
I do remember I used to work for geographical interfaces, where we needed most of the location information. It was a petroleum extraction client. So we needed the location and everything. For that, we used libraries that were more focused on getting the basic distance between two geographical locations and everything. Those kinds of functions, if we wanted to implement them, we used to extract that particular thing from the external libraries and use it.
What is most valuable?
Definitely. Even though you are paying a huge amount, the quality and the time it takes to process big data is definitely lesser. Things are more systematic. All transformations are connected to a pipeline. If you have implemented those things in production and there is any bug, people can easily debug things and identify where exactly the pipeline has an issue. You have easily accessible logs and notifications available. There are a variety of components which can be used to implement any of your logic.
With my current client, I am working for a client who has used Stitch. They have provided it in their environment only. I do not remember from where they have purchased it.
What needs improvement?
During the processing of transformation, sometimes we used to bring up our own logic, like removing duplicates from the data if I have checked two or three keys. Rather than picking up the one which was provided, there is not a process where, for example, if I have four duplicate keys in the record and I have checked two or three fields and ideally we want something which should also utilize the way the record came. The first record which came should not be deleted, and anything other than that should be deleted. Those kinds of functionalities are not provided. It randomly deletes whatever if it finds a duplicate. These kinds of small things, I remember when I worked, I faced challenges because I was doing a migration and my data was not matching. Then I realized these kinds of features are available in this. That is the reason, it is not the problem for Stitch, it is with other ETL tools also.
For how long have I used the solution?
What do I think about the stability of the solution?
It is very rare. I have seen and faced the issue, but it was very rare that there was a network issue, meaning some kind of thing which has broken down and we were stuck and were not able to work on that.
What do I think about the scalability of the solution?
It depends upon if you are using the cloud version, then definitely it is scalable for sure. Or if you are using a normal legacy one, then you cannot scale it that much.
How are customer service and support?
Definitely positive. The kind of support Stitch provides is good.
Which solution did I use previously and why did I switch?
I used to work with Azure back then.
What was our ROI?
Anything which reduces your effort of writing code is beneficial because it will reduce your time. Efficiency is important, and since it is pre-written or reusable code which we are writing, it is going to be more efficient, definitely. If they are providing automation, then definitely it will be beneficial for any of the business or any of the domains you are working on.
What's my experience with pricing, setup cost, and licensing?
To be frank enough, I will say right now that since we are not using it, they are too expensive. I will be very frank to you. If you are using any ETL tool, they are too expensive. People, if they are working on any of the databases such as Snowflake or Databricks, are trying to get involved in using their functionality even more rather than picking up this additional ETL tool and moving towards that.
Which other solutions did I evaluate?
What other advice do I have?
I have moved my role, actually. So basically right now I am into management mostly. But my projects which I am handling right now do have those things that have been handled. Greenplum I have used long back in my previous organization.
If we talk about Greenplum, I worked on that around seven to eight years back. That was a long time ago. PostgreSQL is used frequently by any project, whichever you pick up. PostgreSQL is used very frequently. That is why I mention it.
I have worked with Stitch, Alteryx, Informatica, and DataStage. There are most of the ETL tools I have worked on. I worked on Stitch for five years, two years in DataStage, and I am also certified in Informatica, which is another ETL tool. I also worked for one or two years with Alteryx, which is another ETL tool.
My whole experience is on ETL only. Even though I am doing project management right now, all my products are related to migrations and everything. They involve those ETL things. Since everyone is moving towards Snowflake or Databricks, they are migrating, but the projects still use ETL tools.
I worked as a partner integrator, mostly for data transformations, extraction, and transformation. Transforming the business logics and everything.
Stitch is totally capable of doing those things. It is the way you implement it. It is totally the way you want to customize your code and you want to implement your business logic. If the developer is good enough, they can utilize that tool in a very good way.
Right now, since the way things are moving in the market, everything is on cloud right now. Mostly, even the current projects I am working on, everyone is migrating towards cloud.
The learning curve is moderate. It is not that complicated.
You need a skilled resource because what happens is the person who is a coder, such as I am a coder in Python and JavaScript, can easily go and start it. But when you are working with such kinds of tools, you need a person who is tool-specifically skilled. Have they worked on this particular tool? They need a person who has some knowledge about the tool. They should know where which functionality is available. Those kinds of details they should know.
I give this review an overall rating of eight.
Which deployment model are you using for this solution?
Other
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Other