Stitch is primarily used for integration between multiple systems in our environment. On top of that, we have data integration and data analysis. We draw information from multiple systems, both systems that we manage ourselves, and we have a number of data partners that manage their data. We hook into their environments, use their information and our information, mix it all up, and then develop new products. Primarily, we're using it for analysis around the data that's captured around the environment. Our council is responsible for looking after a large chunk of New Zealand and monitoring air and water and sea, and managing the quality of a large number of different environmental factors. We need to keep reporting that in real-time aggregation on a monthly and yearly basis. We're looking at how climate change is affecting our environment and reporting on anything about the environment.

The benefits of Stitch that we see are the ease of being able to integrate and visualize data that's in other systems. We're dealing with APIs, both XML and JSON, and pretty much every integration format that you can imagine. The platform itself is really easy to hook into all of those different outside sources and bring it in and transform it into a format that meets all of our internal systems. It's just the ease of use really.

We're connecting to everything with the extensive connector library in Stitch. There are pre-built connectors already in there, but one of the big powerful things about Stitch is that a lot of these specialized systems aren't standard connectors, like SAP or any of those big systems. Because the components themselves are very configurable, we can actually connect to systems that don't have a standard interface connection or have an API but the XML format isn't quite normal or the JSON format isn't quite normal or whatever format they're in. Stitch allows us to connect to those specialized data sources and transform them into a format that's easy for us to deal with internally.

The transformation language feature helps with data consistency in our case, as we convert everything into JSON. JSON is our internal standard for data formats. Stitch makes that whole data transformation part really easy. That was the biggest selling point for us.

The schema evolution capability helps to maintain data integrity because a lot of these systems don't change, but when they do change, we need to be able to manage that change in a controlled manner. The ability to transform based on different versions is where Stitch is really powerful. We can set up version one using a transformation standard, but when we upgrade to version two, it transforms the whole mapping capability. It makes it really easy to go from version one to version two in real-time. We don't have to convert the whole database; we can transform it as we need it, which means that we don't have to go through a big migration anytime there's an upgrade or change from the source system.

One of the key things is that we need benefits from real-time insights and monitoring tools in the product because we need to know when something goes wrong or something isn't part of the standard pattern. The jobs themselves alert us when something goes a little bit differently or doesn't fit the existing pattern. It doesn't stop the whole job, but it gives us an alert and makes us aware of what's actually going on so that we can intervene if we actually need to. But a lot of the time it's more of an information notification about what has changed rather than a critical alert. However, over time, those little variations do grow and grow, and the platform allows us to actually keep track of how far things are actually growing or changing so that we can then interpret whether it's the best time to actually make a change to the whole program of work, or whether we can just live with it for a bit longer.

The automation features in the product have a significant impact on our data management process because we would not be able to do what we currently do if we didn't automate. The number of data sources that we're dealing with is growing, and every year there's probably twice as many as there were the previous year. We've only got a small team, so by automating, we don't have to spend a lot of time building new internal resources to manage it. We've pretty much got the same size team as we did four years ago. The automation just makes our lives really easy.

The best thing that we've seen coming out of Stitch is the Talend and Qlik relationship becoming more of a single entity rather than two entities. A couple of years ago there was a distinction between Stitch and Qlik. The biggest challenge when we were dealing with both was that we felt like we were dealing with two different companies. They are quickly merging the two different entities into a single one, and by around the end of this year, they're planning to basically consolidate everything. Until then, it does get a little bit complicated when we have Qlik and Stitch components together. It's not quite exactly one solution. That's probably the biggest negative at the moment. Each month we're seeing improvements in the cloud platform, and they're coming out with capabilities that we haven't even thought about that are going to make our lives easier. Probably the biggest negative they have is they're doing so much good stuff, but we don't know about it. It just appears and then we figure it out ourselves, but the communication of changes could be improved. However, the positives are that they're changing so much at the moment.

We've been using Stitch for about three or four years now.

The only real glitches we've experienced with Stitch are when we have uncontrolled change in our environment. All environments are reliant on their firewalls and networks. If we don't change anything to do with the firewall settings, then the last time we had a server issue was probably around seven or eight months ago. It's all running in a cluster, and we can auto-restart servers whenever we feel appropriate for load balancing or the distribution of job resources. Stitch is really stable.

We've found that the scalability of Stitch is really easy. The easiest part is that we just spin up a new server and add it into a cluster, and then it pretty much manages the load balancing across all the servers in the cluster. We keep track of how much CPU is being used at different peak times and all of that. Because we manage most of the load ourselves internally and it's based on internal systems, we're actually pulling more data into our environment and feeding it out to external customers. We can control that load quite well. When it does hit the limits, we just add a new server into the mix and let Stitch manage itself. It seems to work really effectively.

The initial support people from Qlik are really good. The best skill set they've got is that they know when the issue is outside of their knowledge, and they escalate really quickly so that we get to the right people when we need them. We've got a couple of solutions internally that seem a little bit unnecessary, where we ask a question of the support desk and they'll spend a week or two trying to resolve the issue when they really don't have the technical know-how. That's one of the really good things with Stitch support. If they don't know it, they escalate it and we get somebody who can resolve the issue really quickly. The support people are excellent.

The deploying and initial setup of the product is straightforward, but there's a little bit of configuration involved with the remote engines and the runtime environment. The documentation is really good. You quite literally just follow the steps that they say and then double-click an executable and you're away. We have had a couple of issues and they've primarily been our problem rather than the Stitch environment. When we've done upgrades and haven't kept up with the regular monthly updates, we've had things get out of sync. We've had to reinstall the Studio type environment, and it was a complete mess. However, twenty minutes later, we reinstalled it and everything was completely up and running with no problems at all.

We've got a project at the moment that we estimated the integration was going to be around $200,000 to $300,000, and we've been able to achieve the integration for less than a tenth of that, doing it in-house using Stitch. Additionally, long-term, the support and maintenance are better because we don't have to manage bespoke code and we don't have to manage the life cycle around that code. It's all built inside of Stitch and managed for us. It's quite literally a configuration rather than a writing code scenario. We've done a huge saving and that's only one of a number of projects that we've got on where the same sort of things are happening.

The worst thing about the price and licensing model of the product is that it sounds really expensive when you start to go into it because a lot of systems require ten or twelve seats. The challenge with Stitch and the whole Qlik environment is that you don't need ten seats, you just need three or four. The three or four seats will actually do the same as what you would have ten or twelve people in another system. People look at the price of the seats and think it's really expensive and that they're going to have to put ten or twelve people onto it. Actually, if you put ten or twelve people onto it, they'll be running out of work pretty quick because you'll be able to achieve so much more faster and don't need quite as big a team. The cost of the seats is actually cheaper by the amount of value that you're adding to the business. The seats sound expensive, but long-term, they're probably a lot more cost-effective for the amount of volume of work that you're getting through and the value you're adding to the business.

Stitch is definitely worth going for. A little bit hard to provide some metrics or examples of the impact on the data management process, but we have seen the number of transactions per minute increase by about 400 to 500 percent in the last two years. The number of meters that we have out in the real world actually measuring things like air quality and water flow has increased by about 300 percent. The number of databases that we've connected to has probably increased about four times from what we had about three years ago. Everything is pretty much three times as much now as what it was a couple of years ago. Our overall rating for Stitch is nine out of ten.