To support real time analytics, we have developed a smart computing platform to run highly complex algorithms at massive scale.
- Very large volumes of data being generated rapidly: E.g. retail client generates up to 1.5TB data per day which needs analytics and insight generation
- Highly complex nature of attribution problem: Machine learning, unstructured text processing, econometric models, artificial intelligence based algorithms
- Highly dynamic media environment which requires agile day to day decision making
Kvantum is using GPUs to scale up processing in order to support real time big data computing that leads to significant reduction in algorithms run time & cost. Especially beneficial for our simulation engines, where we load data once and then run billions of operations on the data.