Unifying data lifecycle crucial for AI, says Cloudera

A recent survey, titled "Data Architecture and Strategy in the AI Era", released by Cloudera has highlighted the considerable impact the rise of AI has on data strategies. Notably, the research discovered that 90% of IT leaders believe unifying the data lifecycle on a single platform is crucial for analytics and AI.

This research surveyed 600 data leaders and top IT decision-makers to identify the primary hurdles AI implementation faces. The leading challenges mentioned were the quality and availability of data (36%), scalability and deployment issues (36%), and integration with existing systems (35%).

The report unveiled three essential prerequisites for effective AI: a modern data architecture in business strategy, which benefits from streamlining data analytics and gaining flexibility in handling different data types; unified data management, where flexible and scalable cloud management technologies are necessary to transform information into insights; and versatile and secure data platforms, where embracing a multi-cloud or hybrid capabilities is vital for an organisation's adaptability to change.

According to Abhas Ricky, Chief Strategy Officer at Cloudera: "As more enterprises look to transform their businesses to build digital and AI-ready solutions for their customers, they are choosing a hybrid and multi-cloud strategy, which in turn creates 'data sprawl and architectural overruns' across LOBs, functional units, business applications and practitioner teams. In order for them to effectively leverage AI capabilities, organisations need to design and embed standardized, use case-centric data architectures and platforms that will allow disparate teams to tap into all of their data – no matter where it resides - whether on-premises or in the cloud."

Additional key findings from the survey revealed that a modern data architecture, essentially grounded in business strategy, simplifies data analytics processes (40%) and enhances flexibility in handling all data types (38%). In regards to end-to-end data management necessary for AI model development, the main hindrances as per the survey were the volume and complexity of data (62%), data security (56%), and governance and compliance (52%). Accepting a hybrid data management approach using both on-premises and public cloud deployments is vital in the current analytical strategy - 93% of participants concur that "multi-cloud/hybrid capabilities for data and analytics are key for an organisation to adapt."

Abhas Ricky stressed: "At its core, enterprises want to achieve top-line results from their data strategy and supercharge their AI initiatives at a price point that is not prohibitive to the bottom line. Organisations that are looking to get the most out of their data need to rapidly build and deploy a modern platform and AI architectures that support that mission. Cloudera is committed to helping customers tackle their toughest data and AI challenges as the industry's only hybrid, multi-cloud data platform for data anywhere."

The respondents in the survey, carried out by Foundry Media for Cloudera, included over 600 IT decision-makers from the U.S., Europe, and APAC. The research targeted firms with annual revenue of more than $500 million or more than 1,000 employees globally. All participants were data and IT decision-makers with titles of director and above (or equivalent) that play a significant role in the selection of data-related products and services.