While the gains in efficiency and automation provided by AI and machine learning are transforming a wide swath of business processes, the proliferation of enterprise AI is itself driving changes in enterprise IT infrastructure. One notable development is the emergence of the cloud-based AI platform, an all-in-one hosted system for developing and deploying machine learning models. According to new data from 451 Research’s Voice of the Enterprise: AI and Machine Learning Infrastructure 2019 survey, 97% of adoptees of enterprise AI plan to use cloud-based AI platforms, and 89% of these companies say they will increase budgets for these services over the coming year. These numbers align with the value of these tools, which provide users with many important features – such as flexibility, scalability and technical capabilities – that help to overcome the major barriers to adoption of enterprise AI.

The 451 Take

Given the increased prevalence of cloud infrastructure and the growing adoption of machine learning applications in the enterprise, it’s no surprise that these trends have converged in the form of cloud-based AI platforms. These platforms have several qualities – such as scalability and modularity – that accelerate the adoption of enterprise AI by making it easier to build and train machine learning models. Given their incumbent advantage in the cloud market and their record of innovation in the machine learning space, it makes sense that the major public cloud vendors – Microsoft, Amazon and Google – are the top purveyors of cloud-based AI platforms. In the future, 451 Research expects these and other vendors to continue to augment the capabilities of their services with faster infrastructure, better machine learning components and other features to address emerging issues in this market, including model explainability and algorithmic bias.


What are 'Cloud-based AI Platforms' and Why are Companies Adopting Them?

Cloud-based AI platforms facilitate the development and deployment of machine learning models and applications in the cloud. They share the following key components:
  • Ability to upload proprietary data and pre-built models
  • Workspace where users can train new models
  • Access to modular pieces, such as pretrained models
  • Tools to deploy, manage or monitor models
  • Hosted or partially hosted infrastructure
There are many reasons why enterprises would seek to adopt a cloud-based AI platform. As any machine learning practitioner or data scientist would tell you, building a machine learning model is complex and complicated. First, it requires specific skills and knowledge. Second, it necessitates a lot of curated data and compute resources. Finally, the process itself is iterative and experimental – it is as much an art as a science.

The aforementioned components of cloud-based AI platforms help to alleviate these pain points. Pretrained models enable experts to construct machine learning pipelines more quickly and allow nonexperts to dip their toes in the water. Those who want to use their own data to build and train custom models can do so using these platforms. The hosted arrangement eliminates the need for adoptees to buy and maintain expensive hardware; instead, rented resources run the intensive workloads when needed. Finally, a series of additional tools and features help to further abstract the complexity, making building and deploying models and applications significantly easier.

Figure 1: Adoption of Cloud-Based AI Platform

Source: 451 Research Voice of the Enterprise: AI and Machine Learning 2019

These benefits have caught the notice of those looking to adopt enterprise AI. As the figure above demonstrates, cloud-based AI platforms are a growing component of enterprise AI infrastructure. Nearly half (49% ) of enterprises that have already implemented an AI initiative or plan to within the year currently use these services, and a whopping 97% plan to within three years’ time. Furthermore, 89% of companies planning to adopt cloud-based AI platforms say they will increase budgets for these services over the coming year.

Major Vendors of Cloud-based Platforms

Not surprisingly, the data shows the dominant vendors in this space to be the major public cloud vendors, with Microsoft having a slight advantage. AWS, Azure and GCP have all invested heavily in building out their portfolio of cloud-based tools and services, and GCP in particular positions these capabilities as a differentiator as it tries to build its market share. Although its public cloud offering has faltered, IBM maintains a stable of machine learning assets under its Watson umbrella.

The other vendors in this space include the large business software vendors such as Oracle, Salesforce, SAP and SAS. These organizations are already used to storing and analyzing enterprise data, which can then readily be combined with machine learning services to get enterprise AI projects jump-started. Finally, an emerging class of vendors includes niche data science and machine-learning-specific vendors such as DataRobot, BigML and

What Might the Future Hold for Cloud-based Platforms?

These data points paint a dramatic picture of the what the future of enterprise AI will look like. First, cloud-based AI platforms will be a pervasive feature of enterprises AI strategies. This fact doesn’t necessarily mean that on-premises options will cease to exist. There are still many reasons for enterprises to use their own infrastructure, such as the security of AI systems or data privacy. But cloud-based platforms will have a ubiquitous presence and, at a minimum, will play an important role in the process of building and deploying enterprise AI.

Second, given the uptake of their products, the purveyors of these platforms – especially the dominant public cloud vendors – will wield enormous influence over the direction the technology goes. These providers will prioritize, for example, which industry use cases of AI and machine learning to develop. They’ll also be charged with building out important features – namely explainability and anti-bias tools – needed for the next generation of enterprise AI. Finally, these companies will oversee the security of complex and integrated series of software systems comprising data models driven by their customers’ sensitive data.
Nick Patience
Founder & Research Vice President

Nick Patience is 451 Research’s lead analyst for AI and machine learning, an area he has been researching since 2001. He is part of the company’s Data, AI & Analytics research channel but also works across the entire research team to uncover and understand use cases for machine learning. Nick is also a member of 451 Research’s Center of Excellence for Quantum Technologies.

Jeremy Korn
Senior Research Associate

Jeremy Korn is a Senior Research Associate at 451 Research. He graduated from Brown University with a BA in Biology and East Asian Studies and received a MA in East Asian Studies from Harvard University, where he employed quantitative and qualitative methodologies to study the Chinese film industry.

Keith Dawson
Principal Analyst

Keith Dawson is a principal analyst in 451 Research's Customer Experience & Commerce practice, primarily covering marketing technology. Keith has been covering the intersection of communications and enterprise software for 25 years, mainly looking at how to influence and optimize the customer experience.

Want to read more? Request a trial now.