What Lies Ahead for Big Data and Analytics in 2018

As we enter a new year new trends will take their turn in the spotlight. This is no different for the big data landscape.

We have come a long way since the term "big data" swept the business world off its feet as the next frontier for innovation, competition, and productivity. Hadoop, NoSQL, and Spark have become members of the enterprise IT landscape, data lakes have evolved as a real strategy and migration to the cloud has accelerated across service and deployment models.

On the road ahead, the demand for real-time analytics will continue to skyrocket alongside growth in IoT, machine learning, cognitive applications, and data governance.

DBTA recently held a special roundtable webinar featuring Paul Nelson innovation lead, Accenture Analytics, Leena Joshi, VP of product marketing, Redis Labs, and Balaji Mohanam, senior product manager, Qubole who discussed upcoming challenges and opportunities in 2018.

From data lakes to business outcomes, according to Nelson, business owners demand outcomes and just having the lake is not enough.  Data Science must produce results because play and exploration are not enough. Everyone wants real-time, day or week-long analysis is too slow and users need immediate actionable outcomes. Everyone wants Artificial Intelligence, systems must be secure, and users want machine learning and IoT as massive scale.

Joshi said Redis Labs can assist enterprises who are looking to gain real-time insights as the solution offers performance, simplicity, and extensibility.

Three transformations are disrupting data processes, Mohanam said. These include the transition from the data warehouse to a data lake, the cloud, and machine learning along with deep learning and AI.

A prescription for success when moving to the cloud includes:


  • Best machine configuration for the workload
  • Best Engine for the workload
  • On demand and elastic; automatically scale up
  • or down


  • Initial provisioning in min/hours, not months
  • Change configurations dynamically
  • Compute and Storage scale independently


  • Pay only for what you actually use
  • Use spot instances to reduce cost by up to 80%

An archived on-demand replay of this webinar is available here.