We would like to keep you up to date with the latest news from Digitalisation World by sending you push notifications.
Key findings include:
•Maturing Market: Investment in AI/ML is continuing to rise, with 86% of survey respondents saying their organisations have increased their yearly AI/ML budget from 2020 to 2021. The same percentage (86) said their organisations prioritise AI/ML above other IT initiatives, with 42% placing AI/ML as their leading IT priority.
•Operational Issues Delay ROI: The research found that organisations are running into increasingly complex post-deployment operational issues, with 87% of survey respondents struggling with long model deployment timelines. The majority, 59% of surveyed organisations, require at least one month to deploy a trained model to production.
•Infrastructure Complexity: Every enterprise has a growing, diverse and often disconnected combination of infrastructure, tooling and specific use cases and requirements for AI/ML. The majority of surveyed organisations (64%) deploy models and data to support more than 10 regions across the globe, while 22% need to support more than 20 regions worldwide. Additionally, 37% of respondents have AI deployed across a hybrid environment for model deployment, combining on-premises infrastructure with cloud environments, while 28% have a multi-cloud environment.
•Performance, Compliance and Security Implications: 85% of respondents are struggling with IT governance, compliance and auditability requirements related to their AI/ML deployments, and 25%—the largest percentage for any single challenge—named IT security their top AI/ML challenge. Most companies (83%) said they have SLA requirements for model latency and significant geographic distribution can present performance challenges.
“AI is enabling the most competitive companies to revolutionise their business operations and disrupt decades-old industries,” said Executive Vice President of MLOps at DataRobot, Diego Oppenheimer. “However, organisations are running into more complex operational concerns: corporate governance, IT security, risk management and multinational regulation. An end-to-end AI/ML platform with enterprise-grade machine learning operations is the only way to manage this growing complexity and maximise business impact.”
