The AI capabilities of Google Cloud Platform

By Chris Cox, Technical Lead, Google Cloud Platform, CTS.

  • 5 years ago Posted in

According to a survey of 51 senior executives, 85 percent of UK organisations are gearing up to make greater investments in artificial intelligence (AI) and machine learning over the next three years.


Deloitte’s first Digital Disruption Index Report found that half of the respondents were planning to invest more than ?10 million into various digital technologies by the year 2020, including increased investment in AI, cloud platforms, and IoT.


Through a series of questions designed to track the uptake and impact of new technologies across public and private sector organisations, Deloitte also found that more executives than ever now see a growing role for AI in their businesses.


AI, machine learning and cloud computing were identified by business leaders as technologies likely to have the biggest impact in the coming years. However, despite their belief in the future importance of AI, Deloitte found that just 22 percent of businesses had already committed spending towards it, and two thirds of those expected to spend under ?1 million on AI technology during 2017.


While investment remains low and organisations test the waters with AI pilot schemes, the results show that more leaders than ever are looking to the advantages that AI can bring to their business.


Google’s Machine Learning Platform


Google Cloud Platform (GCP) represents one of the most easily accessible paths for those looking to take their first steps towards integrating AI into their business.


Until recently, the type of machine learning provided by GCP was used almost exclusively by large data-rich organisations like Facebook, Microsoft and IBM, or by bodies such as research institutes and the medical field which are typically in need of powerful analysis tools.


Over the past few years, however, Google has been busy seeking out industry expertise and building on its existing cloud platform to put the power of machine learning in the hands of enterprise customers.


After joining Google last November, leading machine intelligence researcher Fei-Fei Li promised to use Google’s cloud technology to lower AI’s barrier to entry. Since making that pledge, Google has launched a range of machine learning products designed to help businesses to take advantage of the powerful cloud platform.


At the core of Google’s AI offering is the Machine Learning Engine - a service that enables businesses to build their own scalable models for processing, predicting and analysing thousands of users and terabytes of data.


Supported by the TensorFlow framework and Google’s TPU hardware, the Machine Learning Engine gives businesses access to a scalable and ‘teachable’ tool to support data processing and other business applications, all handled by the powerful Google Cloud for faster training times and improved performance and accuracy.



Powerful APIs For Business


Beyond the raw technical power of the Google Cloud Platform, utilising machine learning and AI opens up access to several different APIs designed specifically to support businesses applications.


For starters, Google’s Dialogflow Enterprise Edition gives businesses the capability to build communication tools like chatbots for websites and messaging applications. Thanks to a strong machine learning foundation, Google claims that GCP-based chatbots are more capable of recognising context and intent during conversation with users, and can provide more accurate and natural responses than ever before.


Machine learning also enables businesses to make use of Google’s powerful translation and speech recognition APIs. Both of these services give users more ways to interact with a business through application voice control, and provide automatic transcription and translation of any conversation - a benefit for users and for businesses.


Any data collected by an organisation using GCP can also be subjected to advanced analysis. Google’s Natural Language API can be used to determine the meaning of text, and can even extract information about people, events and places from any passage of text it is shown. It can also be used to automate the analysis of social media posts, messaging applications and web chat apps - understanding the sentiment of customer reviews and determining customer intent from text logs.


Image and video data can be analysed too, with features like facial recognition, image classification, and text detection dramatically increasing the speed at which different content types can be analysed and put to use by a business. The potential uses of this technology range from visual diagnosis for the healthcare industry, through to automated vehicle control and facial recognition for security purposes.



The Benefits of AI Adoption


Aside from the uses for individual APIs, the powerful data analysis made possible by machine learning has many proposed applications. From optimised delivery routes for logistics businesses to personalised advertising, enhanced credit checking, customer processing and retail supply and demand prediction, AI has the potential to revolutionise the way many industries do business.


As more and more business leaders move to make investments in AI over the coming years, as discovered by Deloitte’s survey, many will inevitably look to the Google Cloud. Backed by powerful infrastructure and supported by flexible and business-focused APIs, GCP is an obvious choice for any business taking their first steps towards AI integration.

By Rosemary Thomas, Senior Technical Researcher, AI Labs, Version 1.
By Ram Chakravarti, chief technology officer, BMC Software.
Anders Brejner, Investment Director and Enabling Solutions Lead at Circularity Capital, discusses...
By Andy Baillie, VP, UK&I at Semarchy.