Case Study:
Big Data Provides Big Results for Hyundai


Pressured by new dynamics and a changing competitive landscape, the customer wanted to implement the use of real-time telemetry but had little idea of how to do so given the particularly challenging big data requirements, i.e. the global reach of the customer’s product—a high volume of cars, each with hundreds of sensors producing gigabytes of data per hour of operation, and its limited knowledge of cloud computing, a fundamental requirement for capturing big data. They wanted not only a solutions provider, but also a strategic partner to help them understand and navigate the complex world of cloud computing.

The Solution

Unitas Global (Solinea) began the engagement by bringing together key stakeholders—both representatives from business units and technology divisions—to devise a cloud strategy in line with the company’s business objectives.

Among other tasks, we optimized a set of cloud offerings, determined the most efficient workloads, recommended an appropriate target platform (in this case a hybrid cloud), performed a competitive assessment, developed an economic model in order to estimate the value of the of new offerings, we assessed the existing technical infrastructure, defined gaps, devised an implementation roadmap, and then ran a proof of concept which was extremely successful.

In the POC, we compared the company’s legacy big data appliance to the new cloud model via the following steps:

  • Loading a predefined and parsed data set on Hadoop
  • Map/Reduce transforms the data to key and value pairs
  • Submitting the Map/Reduce job
  • Monitoring the job for completion

A prescriptive architecture as code approach was taken, as well as shared services and an open community between development groups. This streamlined workflow around a modern CI/CD tool chain, while at the same time enabled code reuse across development teams. This greatly reduced duplicate efforts that were discovered during the application review stages.


From the POC, we were able to prove the value of commodity big data. The client subsequently engaged us to define the technical cloud architecture to support the solution, the basis of which was:

  • Commodity servers for compute and storage nodes
  • Private Cloud infrastructure as a service
  • Apache Hadoop
  • Custom code


  • Faster time to market for new services and application changes
  • Deployment shortened from 100 days to 12 hours
  • Optimized infrastructure spend by relying on AWS for bursting services during peak demand
  • More time for developers to spend on building applications rather than worry about infrastructure 

Contact us for more information: