Unlike traditional analytics and business intelligence applications, the proposed solution will tackle two pains at the same time: the first is the current limitations of retaining and managing the accumulated massive data streams and associated users' footprint on the operation value chain. The second is unleashing value out of the huge amount of data collected over months and years of operations. We are taking on profile of data ingestion of dozens TB per hour and date scan rate of hundreds GB per second and even can grow linearly beyond this level.
For the first pain, the solution will provide scalable open big data platform implementing the technologies proved by giants like Google, Yahoo, LinkedIn and Facebook. The platform will enable very fast processing of data and ability to ingest data from multiple sources. Longer data retention will be facilitated by cost-optimized computing and storage units while the cloud-nature of the solution will enable virtually unlimited growth of the platform. The platform will also be hosting multiple technologies working together to address the second pain point of unleashing the value out of the data. Technologies like Massively Parallel Processing, In-Database Machine Learning, Stochastic Arrays, Resilient Services, Container-Based Microservices and Cloud Foundry will be integrated and employed in the proposed platform.
On top of the proposed Big Data platform, data driven models will be developed and operated to extract the business value out of the data. Thus converting the data from liability to asset and make use of the gold mine of customer-intensive data to develop new business opportunities, offer new services and grow the revenue. The possibilities with such data driven models are unlimited where a an Agile product life cycle will be put in place to ensure faster results and continuous development & enhancement. Data Sciences techniques will be utilized to develop models for contact center, security, marketing planning, network operations enhancements, event driven promotions, data monetization and further more. Further more, these models should be able to provide real-time value by utilizing the power of processing provided by the big data platform part of the solution.