Putting AI and Machine Learning Workloads in the Cloud

Artificial Intelligence (AI) And the Machine Learning (ML) are some of the most Ultra Enterprise Technologies It has captured the imagination of councils, with the promise of efficiencies and lower costs, and the public, with developments such as self-driving cars and autonomous quadcopter air taxis.

Of course, the reality is more realistic, as companies look to AI to automate areas such as online product recommendations or spot flaws in product lines. Organizations are using AI in vertical industries, such as financial services, retail, and energy, where applications include fraud prevention and business performance analysis for loans, forecasting demand for seasonal products and processing massive amounts of data to improve power grids.

All this falls short of the idea of ​​artificial intelligence as an intelligent machine along the lines of 2001: A Space Odyssey Hal. But it’s still a fast-growing market, driven by companies trying to get more value from their data, and Business Intelligence and Analytics Automation To improve the decision-making process.

For example, industry analyst firm Gartner expects the global market for artificial intelligence software to reach $62 billion this year, with the fastest growth coming from knowledge management. According to the company, 48% of CIOs surveyed have deployed AI and machine learning or plan to do so within the next 12 months.

Most of this growth is driven by developments in cloud computing, where businesses can benefit from Low initial costs and scalability of cloud infrastructure. For example, Gartner cites cloud computing as one of the five factors driving the growth of AI and machine learning, as it allows companies to “experience and run AI faster with less complexity.”

In addition, large public cloud providers It develops its own AI modules, including image recognition, document processing, and advanced applications to support industrial and distribution operations.

Some of the fastest growing applications of artificial intelligence and machine learning revolve around e-commerce and advertising, as companies look to analyze spending patterns, make recommendations, and use automation to target ads. This takes advantage of the increasing volume of business data already in the cloud, reducing the costs and complexity associated with data transmission.

The cloud also allows organizations to take advantage of advanced analytics and computing facilities, which are often not cost-effective to build in-house. This includes the use of dedicated graphics processing units (GPUs) and very large volumes thanks to cloud storage.

“These capabilities are beyond the reach of many of the on-premises offerings of many enterprises, such as GPU processing. This illustrates the importance of cloud power in enterprise digital strategies, Lee Howells, head of AI at PA Consulting, explains.

Companies are also building expertise in their use of AI through cloud-based services. growth area is AIOpswhere organizations use artificial intelligence to improve their IT operations, especially in the cloud.

else MLOps, which Gartner says is running multiple AI models, creating “structured AI environments.” This allows companies to build more comprehensive and functional models from smaller building blocks. These blocks can be hosted on on-premises, on-premises systems, or in mixed environments.

AI offerings for cloud service providers

Just as cloud service providers provide the building blocks of information technology — computing, storage, and networking — they are building a range of artificial intelligence and machine learning models. They also provide services based on artificial intelligence and machine learning that third-party technology companies or companies can integrate into their applications.

These AI offerings don’t need to be end-to-end operations, and often don’t. Instead, they provide jobs that would be costly or complex for the company to provide for itself. But they are also functions that can be performed without compromising company security or regulatory requirements, or that involve large-scale data migration.

Examples of these AI modules include image processing, image recognition, document processing, analysis, and translation.

“We work within an ecosystem. We buy bricks from people and then build homes and other things out of that brick. Then we deliver those homes to individual customers,” said Mica Vaino Mattella, CEO of Digital Workforce, a Robotic Process Automation (RPA) a company. The company uses cloud technologies to expand the provision of automation services to its customers, including a “robot as a service,” which can run either on Microsoft Azure or a private cloud.

Vainio-Mattila says AI is already an important part of business automation. “Which is probably the most prevalent is Intelligent document processingwhich gives essential meaning to unstructured documents.”

“The goal is to make these documents meaningful to ‘bots,’ or automated digital agents, who then do things with the data in those documents. This is where we’ve seen the most use of AI tools and techniques, and where we’ve applied AI ourselves the most.”

He sees a growing push from large public cloud companies to provide AI tools and models. Initially, this was for third-party software or service providers like his company, but he expects cloud solution providers (CSPs) to deliver more AI technologies directly to user companies as well.

“It’s an interesting space because the big cloud providers — obviously led by Google, but closely followed by Microsoft and Amazon, and others, IBM as well — have implemented services around ML and AI-based services to decode unstructured information. That includes image recognition classification, or translation.

These are “general purpose” technologies designed so that others can reuse them. Business applications are often very specific to a use case and need experts to adapt them to the needs of a company’s business. The focus is on back office operations rather than applications such as self-driving cars.

Cloud providers also offer “domain-specific” modules, according to Howells of PA Consulting. These have already developed in financial services, manufacturing and health care, he says.

In fact, the range of AI services offered in the cloud is vast and growing. “Great [cloud] “Players now have models that everyone can take and run,” says Tim Bowes, associate director of data engineering at consultancy Dufferin. “Two to three years ago it was all third-party technology, but now they build their own proprietary tools.”

Azure, for example, offers Azure AI, with AI models of vision, speech, language and decision-making that users can access via AI calls. Microsoft divides its offerings into Applied AI Services, Cognitive Services, Machine Learning, and AI Infrastructure.

Google Offers artificial intelligence infrastructureVertex AI, ML platform, data science services, media and speech-to-text translation, to name a few. Cloud Inference API allows companies to work with large data sets stored in the Google cloud. The company, unsurprisingly, provides cloud GPUs.

Amazon Web Services (AWS) also provides a wide range of AI-based services, including image recognition, video analysis, translation, and conversational artificial intelligence for chatbots, natural language processing, and a set of services aimed at developers. AWS is also strengthening its health and industry units.

Large enterprise software and SaaS (SaaS) software providers also have their own AI offerings. This includes Salesforce (ML and predictive analytics), Oracle (ML tools including pre-trained models, computer vision, and NLP), and IBM (Watson Studio and Watson Services). IBM has developed a specific set of AI-based tools to help organizations understand environmental risks.

Specialized companies include H2O.ai, UIPath, Blue Prism, and Snaplogic, although the latter three are best described as intelligent automation or Companies RPA of artificial intelligence providers.

However, it is a fine line. According to Jeremiah Stone, Chief Technology Officer (CTO) at Snaplogic, companies often turn to AI on an experimental basis, even when more mature technology is more appropriate.

“It’s possible that 60% or 70% of the efforts I’ve seen are, at least initially, to start exploring AI and machine learning as a way to solve problems that can be best solved with better understood approaches,” he says. “But that’s tolerable, because, as people, we’ve always been very optimistic about what software and technology can do for us — if we don’t, we won’t move forward.”

He says that experimenting with AI will yield benefits in the long run.

The limits and prospects of cloud-based AI

There are other limitations to AI in the cloud. First and foremost, cloud-based services are best suited for general data or general operations. This allows organizations to overcome the security, privacy, and regulatory hurdles involved data sharing with third parties.

AI tools counter this by not transferring data – it remains in the local business application or database. And security in the cloud is also improving, to the point that more companies want to take advantage of it.

“Some organizations prefer to keep their most sensitive data in the workplace. However, with cloud providers offering industry-leading security capabilities, the reason to do so is dwindling quickly,” says Howells of PA Consulting.

However, some companies prefer to build their own AI models and conduct their own training, despite the cost. If AI is the product – and Driverless cars Prime example – the company will want to own the intellectual property rights in the forms.

But until then, organizations will benefit from areas where they can use public data and models. Weather is one example, image recognition is likely another.

Even companies with very specific requirements for their AI systems may take advantage of the vast data resources in the cloud to train the model. Likely, they may also want to use the synthetic data of cloud service providers, which allows for modular training without the security and privacy concerns of data sharing.

And few in the industry would bet on those services that come, first of all, from cloud service providers.

Leave a Comment