Artificial Intelligence as a Service (AIaaS) on cloud promising next industrial revolution
Even though cloud computing has existed since 2000, It only emerged as one of the leading technologies in 2008 by adapting to existing technologies such as virtualization, service oriented architectures.Today we see most companies migrating to the cloud by moving their services to Amazon, Google, IBM, Microsoft and a variety of other cloud providers.
(Figure 1: AIaaS on Cloud source )
Artificial Intelligence as a Service (AIaaS):
Kevin Kelly a digital visionary on TED says “So we have many ongoing tendencies right now, and I think one of the chiefs among them is this tendency to make things smarter and smarter. I call it cognifying — cognification — also known as artificial intelligence, or AI. And I think that’s going to be one of the most influential developments and trends and directions and drives in our society in the next 20 years” He concludes saying “The most popular AI product in 20 years from now, that everybody uses, has not been invented yet. That means that you’re not late”
Before going over facts on how AIaaS on cloud promising next industrial revolution, I would like to shed some light on basics of AIaaS on Cloud.
(Figure 2: Theory of AIaaS on Cloud Source )
Cloud AI can be categorized into 4 quadrants in the 2×2 matrix. On Y axis we have the type of client involved which can be physical or software application. On another axis, we have the type of data upload which can be structured such web transactions, stock levels, GPS coordinates and other sensor reading. Human beings are capable of processing structured data but tend to become overwhelmed as data becomes large, at this level AI can process functions with relative ease. Unstructured data include audio, digital photo and video all of which in earlier was difficult for computers to make sense of, but today it is possible with cloud AI services can able to make sense of unstructured data and respond.
With this, I would like to go over current updates on companies adopting the cloud AI services in their business and how this has improved their overall productivity
Highlights of Google Cloud Next ’17 – Day 1 Keynote
Diane Greene, SVP of Google Cloud started the keynote with “Cloud is where a lot of digital revelation is going on in IT industries.” Some of the key points from her speech are
- Google Cloud Platform (GCP) inbuilt with AI and ML engines is capable of finding previously hidden patterns in data to predict newer results which are helping in business productivity.
- Datacenters are now equipped with new generation TPUs which is now capable of handling deep learning problems.
- GCP is now having world class infrastructure with AIaaS features
- Michael White CTO, SVP Disney Consumer Product, and Interactive media mentioned that they have moved software development to QA on GCP, and now all their top ranking games run on GCP. With this, as a software development point of view it frees up their developers and resources, which helps their developers to focus on more innovative work and able to take ML mindset on their road to AI, they now strongly believe this will enhance their overall business
- Bernd Leukert, Member of Executive Board SAP says, “SAP has more than 345 thousand enterprise customers, 76 % of the world transacts revenue uses a SAP system. Now “SAP HANA” which is analytical processing tool used by their 14000 customers is now available on GCP, with this SAP HANA has now had features like advanced analytical processing, spatial crafts, predictive capabilities and text analysis which is mainly because of AIaaS inbuilt in GCP”.
- Home Depot yet another firm moved their IT infrastructure on GCP; they are now expecting the retail market to change in next three to five years more than it has since the home depot was founded nearly 40 years ago. They have around 2000 stores in North America which are now running e-commerce infrastructure by making use of AIaas on GCP
- HSBC is one of the large banks in the world established 150 years ago has moved its 93 PB of data onto GCP and running their data analytics tools on GCP. Darryl West group CIO mentioned GCP equipped with AIaaS helping them in important and critical problem solving such
- Anti-Money Laundering (AML): AIaaS APIs are helping to identify AML activity and reduce false positives
- Finance Liquidity Reporting: with AI and ML support on cloud liquidity processing has reduced from 6 hours to 6 mi
- Some other services they are running on cloud are Risk Reporting, Risk Analytics, and Valuation services
- RT Pittman, chief product officer from eBay says “Outsourcing your IT infrastructure and concentrate on your innovation is a strong growth phenomenon”. He says after they moved billions of live listings, 22 years of old data onto GCP, with AI and ML features they are able to build smarter catalogs, able to list the best item for individual at right place right time
Few more cloud companies which are capable with AIaaS are
Amazon AWS cloud is equipped with popular deep learning frameworks to support over million customers in more than 190 countries. Companies such as Expedia, NETFLIX, courser, Unilever running their web application on AWS.
NVIDIA GPU cloud (NGC) is the GPU accelerated platform having the ability to rapidly build, train and deploy neural network models to address complicated AI challenges. Nimbix Cloud Platform is featuring NVIDIA DGX-1TM is a cloud built for applications mainly for deep learning, analytics and HPC applications
Microsoft cloud Azure server equipped with Altera FPGAs is now capable of providing some of the AIaaS
IBM Bluemix cloud platform is yet another cloud with AIaaS with features like blockchain for business, augmenting human intelligence, quantum computing platform, and neural networks etc.
Based on the current updates, it is pretty evident that every cloud provider is trying to bring AI into their cloud, because of the advancement in hardware infrastructure; it can be next generation TPUs, NVIDIA GPU, AWS, Microsoft Azure or IBM Bluemix. Each of this is capable of processing some of the complex AI challenges, and hence every industry are integrating high-performance cloud infrastructure and cutting edge services in their IT environment, this is dramatically boosting the retail market.
So I personally feel bringing AIaaS on cloud is definitely a promising step to bring humans and machines to work together which enables new solutions for the existing problems and also opens up new innovative ideas to build better world