Apps Go Better with AI – IBM, Apple and Coca-Cola

By Charles King, Pund-IT, Inc.  March 21, 2018

One of the most profound corporate IT shifts during the past decade is in developers’ roles as agents of business and technological change. That began with the critical positions developers took in cloud computing and bring your own device (BYOD) adoption. The process accelerated as organizations leveraged their developers to shift focus from massive, monolithic business applications to quickly created, continually evolving apps and mobile services.

As a result, how and how well businesses support developers and their communities have become important points that can eventually impact a company’s market position and ability to compete. While there’s no single, unchanging answer, how vendors work and work together to facilitate developers’ efforts is an important issue to track. The latest evolution of the four-year-old partnership between IBM and Apple certainly qualifies in that regard.

The initial collaboration

Remember that when the pair first announced their budding relationship, it covered fairly conventional ground and was at least partly reflected in partnerships Apple pursued with other vendors, including Cisco. In essence, Apple agreed to work with IBM and its ISV partners to create enterprise-focused apps that would be made available and supported via the Apple Store.

For its part, IBM agreed that along with adopting evangelizing Apple’s platforms and tools to its developer partners, Apple would be IBM’s preferred vendor for endpoint devices (PCs, smart phones and tablets). Alongside that agreement, Apple also agreed that IBM would offer enterprise customers onsite repair and replacement services for Macs, iPhones and iPads.

This all may seem somewhat mundane, but the agreement created a platform for Apple to argue the enterprise readiness and capabilities of its iPhones and other products. The company had been vocal about those issues previously, but it never developed the market position or imprimatur of recognized enterprise-class endpoints, like RIM’s Blackberry handsets. Plus, as consumer demand for smart phones, including premium priced iPhones flattens, it’s increasingly critical for Apple to expand sales among large business customers.

How well have IBM and Apple done together? Not too badly, overall. According to reports, the companies have sold corporate apps to over 800 business customers and built about 100 apps tailored for specific vertical industries. The pair isn’t talking about resulting revenues—not unusual in this kind of deal. But this new announcement signals that the companies are satisfied enough with the deal to keep it rolling.

IBM/Apple – Next moves

While the original partnership focused very much on IBM and Apple, the latest offerings target customers’ developer organizations by delivering services accelerated with artificial intelligence (AI) tools and services. Dubbed Watson Services for Core ML, the new solutions link IBM’s Watson platform with the Core ML developer tools that Apple unveiled last year.

In essence, the solution supplies developers the means for using machine learning to easily and cost-effectively create and deploy AI-enriched business apps. The machine learning heavy lifting will be done behind the scenes on IBM Watson. The machine learning-based models can then be deployed on iPhones via Core ML to support specific AI processes, such as visual pattern recognition.

How would this work in real world circumstances? IBM offered an example of how Watson could be trained via machine learning to recognize specific machinery or electrical equipment. That model could support an iPhone app that field technicians would use to scan electrical equipment and detect any damage, unapproved changes or other anomalies. In fact, the IBM and Apple noted that Coca-Cola is currently utilizing Watson Services for Core ML for improving vending machine inspection.

Final analysis

There’s one other platform that supports similar phone-based machine learning models—Google’s TensorFlow Lite. However, Google appears to be supplying the machine learning training data for those applications while Watson Services for Core ML allows customers to use their own data resources. As a result, IBM and Apple appear to be well ahead of the competition in terms of delivering highly sophisticated machine learning and AI capabilities across an integrated data center-to-endpoint commercial ecosystem.

Overall, this extension of IBM and Apple’s partnership should result in powerful, sophisticated new business- and industry-focused apps that are created more quickly and operate more seamlessly than alternatives that require data to be laboriously moved to and from data centers for analysis. In other words, if organizations and developers want to create and deliver AI-enabled apps that are ready for enterprise Prime Time, Watson Services for Core ML is the way to go.

© 2018 Pund-IT, Inc. All rights reserved.