The client has multiple technologies and clients across different sectors. Melon developed a software tool that can sit on top of all these different technologies they are using and can act as one multi-platform solution that can manage and report on their robots.
Telecommunication & Software
In April 2017, a global leader providing services in strategy, consulting, digital, technology and operations had acquired a Wales-based information technology and services company, focused on the automation of business processes to strengthen their presence in the artificial intelligence and robotic automation space. The project was kicked off with Melon in September 2016, before the acquisition, as a referral from another client we had been working with for two years back then. By the end of the project, the company was already acquired and we delivered the final product to the new owner.
Because the client has multiple technologies and clients across different sectors, they’ve employed Melon to develop a software tool that can sit on top of all these different technologies they are using and can act as one multi-platform solution that can manage and report on their robots.
The first phase was all about building the web platform to which the client can plug different clients using different tools like UI pattern and Blue Prism. It is called the Unified Automation Platform (UAP). They wanted to implement all their experience and create dashboards that can show what went right and wrong, while aiming for a very easy to use screen regardless of what technology or source has been used. They didn’t want to have to log into different control centers, rather they required to have it all in one place so that their workforce can manage it centrally.
In March 2017, the first version of the tool was live and very well received. They plugged two clients to test it, i.e. they integrated two of the control centers they manage into the platform Melon built.
The client also made use of Melon’s UX and UI services. We’ve got the design team to help us create something that not only works very well, but also looks and feels very good.
The next phase was building a scheduling tool as part of the platform. The first phase was focused on building the control center to start and stop the robots, management of the client platforms and monitoring, so the next phase is about integrating the clients’ calendars.
The final phase was creating self-managed robots so that a person can schedule and relocate a robot. The client is again looking to automate this process so that they can create an intelligent scheduling platform. They have 5 people servicing 10-15 clients processing about 200 000 transactions a month and they hope that as they grow their client portfolio they won’t need to hire more stuff.
The CTO says that their domains are artificial intelligence and robotic automation, however capturing and presenting their requirements to Melon so that we can proceed with software engineering was challenging. He also says that the first phase of the project took longer than they anticipated. “But lessons are learned and we hope this won’t happen again.”
The client wanted to outsource the software development of their new platform but at the same time they were after engaged developers. “I enjoyed very much working with Melon because your guys are professional developers who are very good at communicating issues and suggestions,” CTO says. “You are not scared of saying “no” or “there’s a better way of doing this”, which helped us a lot. You have a lot of expertise that we are learning from. Because we are using a dedicated team since October 2016, they feel like a part of our own team.”
The client is very happy with the results. Indeed, it took longer to go live than we were hoping – an extra month, but the tool has been very well received. Their team says that everybody loves it and it’s got the “wow” factor to it. The service team at the control center, who are running the automation work finds that this platform makes their lives way easier and they are very happy, which is what we were trying to do.