InsightEdge used with success in a major european car manufacturer to provide WLTP results
What do you like best?
Easy to start using the solution to prototype and test new ideas. Simplicity to add new features to the first solution delivered to answer new business needs.
Solving a capacity problem was the first motivation for implementation. Still, the choice of InsightEdge aimed to go much further and initiate a data-centric type transformation of several existing systems, leading to simplification, cost-cutting, and a high level of performance.
Even if single request processing time was not the challenge, we get perfect response time and tried to stress the system up to more than 2000 requests per second without noticing an impact on performance.
What do you dislike?
Knowing we can do a lot in many ways due to all InsightEdge capabilities (connectors, storage styles, ...), it can be tricky to find the best implementation for a given problem. Many will work, but if you need to have real-time top performance, you need to work on it and be ready to understand how the solution is managing memory.
Depending on your requirements and the volume of data stored, you may need a lot of memory in your virtual machines.
What problems are you solving with the product? What benefits have you realized?
We solved a technical capacity problem first by implementing a cache where standard cache solutions were inefficient because too much variability in the requests was leading to very low cache hit performance. Feeding a cache would have required years of calculation and will be terabytes big, which was not realistic.
InsightEdge performance and flexibility permit us to process requests on the fly and to be able to try to correlate them with already calculated ones providing the same results, reducing drastically the need to call the target system (up to 90% in only a week, and progressively increasing towards 100%).
Once the cache was full of data, it is a high-performance data source used for marketing or technical very complex analysis.
The benefits were:
- solving in only 3 months the capacity problem
- being able to suppress code in other systems, using the cache as the reference source of information
- replace empirical semi-manual analysis with exhaustive and automated ones, replacing repetitive and time-consuming tasks for business experts
- transform the information system towards a more agile and efficient data-centric one.
