Nokia is one of the world’s largest mobile device manufacturers. An industry leader, their products have become an integral part of the lives of people everywhere, and their technological and design innovations have made the Nokia brand one of the most recognised in the world.
Seeing an opportunity to improve services and products for their clients, they wanted to harness the data that was regularly collected from their devices. They then wanted to use that data to provide insights into the performance of business initiatives.
Traditional analytics platforms are not suited for analysing this kind of data, which is usually unstructured or semi-structured. Further, the volumes are too large to be handled by regular databases / data warehouses in a cost-effective manner.
Nokia decided to adopt Hadoop to build a data management platform. The first generation platform was locally hosted. The analysis was running in bursts - and when it wasn’t, the systems were under-utilised. To address this, Nokia migrated the platform to the cloud - Amazon Web Services (AWS). A custom second-generation platform was then built, since AWS' Elastic MapReduce (EMR) lacked the required flexibility for enterprise adoption.
“We partnered with ThoughtWorks to get the kind of talent and knowledge required to realise our goal. A key principle when one is building a complex, new system is that you need to be agile enough and humble enough to fail and learn from the failure; a key capability that ThoughtWorks embodies as a leader in the agile software development approach.”
The platform was built with three key objectives: reuse processes, reuse infrastructure, and reuse data.
The ThoughtWorks team was involved in driving the architecture and design for multiple components of the second-generation, cloud-based platform. A complex component of the platform – the Job Management Framework – was managed and optimised to deliver rapidly into production every three months.
Once the base capabilities are built, the next step was to migrate the business analytics applications to the platform. A Data Asset Catalog is also being built to provide data at a higher level of abstraction. This will aid analysts in Data Discovery to build reports on top of data assets, truly democratising access to data.
Thanks to this platform, Nokia was able to make EMR on AWS enterprise ready; the analytics applications built on top of the platform can now focus on the business functionality. There's been an increase in agility, leading to a reduced time to market, which means happy customers.