The Trends Influencing Tomorrow's Data Analytics
Eric Mizell, vice president of global solution engineering at Kinetica, discusses how technologies such as machine learning, deep learning and natural language processing (NLP) will influence the future of data analytics.
1. Need for Algorithmic Economy
Humans and current systems are unable to act on the size, shape, and speed of data from newer sources such as social media, mobile apps, and connected devices. This is where automation of insight discovery at scale by algorithmic approaches such as machine learning, deep learning, and NLP can be utilized.
2. Evolution of Data and Analytics Architecture
According to Mizell, data in motion and at rest, transactional and analytic databases, Cloud and on-premise, AI and BI, and more must co-exist and interoperate for the future hybrid world. Workload-specific and complementary analytic solutions must be combined while paying attention to APIs, programming languages, and connectivity to integrate with existing infrastructure.
Delivering business value demands fresh data and real-time insights. Customers want instant gratification. To make the most of fast-paced business opportunities, real-time analytic pipelines simultaneously ingesting, analyzing, and acting on data to deliver timely insights is required.
Flexible and easy-to-use high performance GPU-based analytic technologies have the potential to modernize the analytic infrastructure while also addressing the compute bottleneck posed by big data and shrinking insight shelf life. CPUs are turning obsolete with only 16 to 32 processing cores per chip while GPUs have thousands.
Organizations need fast and easy self-service access to data and analytics for maximum business value. A secure, scalable, and available data-insight-driven culture requires adoption of flexible analytic technologies that democratize analytics, data science, and ML.
The future will require ways to sense, interpret, and respond to data in motion and rest, in real-time and at scale for which organizations must shape their data and analytics architecture to encompass and procure contextual insights from the massive wave of IoT data to increase productivity. Flexibility and scale are critical for cost-effective IoT data management.