Your Company Is Not Ready For AI
The tech industry is abuzz about artificial intelligence (AI). AI is a leading buzzword for corporate governance (boards and directors) and for companies looking to raise capital. Having an AI strategy is believed to enhance a company’s valuation.
Enterprise employees are also hearing more about AI. Some media headlines have even claimed that AI signals the end of the human workforce and fundamentally alters the future of work. While these statements are wild generalizations, if your company’s industry focuses on highly repetitive, transaction-oriented business, it is ripe for automation.
With the hype around any emerging technology, it is the IT mission to qualify and quantify the AI opportunity in support of the corporate governance model. Even if you are excited about what AI technologies can introduce to your business, not everyone in your organization likely shares the same enthusiasm. And you may, in fact, find that your company is not yet ready for AI.
How do you determine if your organization is ready for AI technologies? A simple true/false question will help clarify:
Is the business comfortable in dealing with ambiguity?
The CIO and IT team may have experience with ambiguous outcomes, such as found when implementing different security schema. The CFO, however, likely deals in very specific terms about schedules and return on investment (ROI). If the business leaders are unable to get past this hurdle, implementing AI technologies in the enterprise should be reprioritized to align with a supportive company culture.
It may seem like AI is a relatively new topic for the enterprise, but numerous case studies and lessons learned are already in the public domain. Some of the best practices distilled from these early learnings include:
- Assign an executive sponsor to AI projects that flourishes in the face of ambiguity and has an eye for uncovering insights and enhancing KPIs.
- Establish clear business goals for pursuing AI technology in your organization
- Create a pilot project in parallel with production systems to understand the impact on workflow and security
- Expect to iterate and tune machine learning algorithms before observing benefits
Enterprises are finding success deploying AI technologies in several ways:
- Data analytics are available now for every business in every industry. Algorithms for common business functions are available as off-the-shelf packages with integration services from the software provider.
- For enterprise security, many commercially available platforms today work off pattern matching (such as malware signatures). This remains a reactive approach to security. AI machine learning enables identification of unusual behavior and usage by establishing a baseline and escalating outliers in real-time.
- You have likely experienced a recommendation engine through e-commerce sites such as Amazon where a product page includes, “Other people that looked at this item also considered…” The same approach can be used across employee roles to suggest how peers responded to action requests from business tools, such as expense approvals or management applications.
These business opportunities for AI technologies apply directly to the demands of enterprise mobility. If you have ever rolled out a new system or service to the mobile workforce and received a tepid response, you understand how challenging it is to align the capabilities of a service with the needs and demands of workers. Machine learning creates an opportunity to personalize the mobile user experience.
Employees can have the flexibility to choose the type of device that is most appropriate for their role, regardless of the system’s support for OS or app. This personalization approach utilizes microservice architecture. If screen size is more important than one-handed mobility, an employee could use a tablet and communicate with line of business applications via email. Alternatively, if mobility is the most important requirement for some roles, perhaps a smartphone that uses instant messaging apps is the preferred method of interaction. No longer is there a prerequisite for a business platform to offer a mobile app. App and service connectors bridge the line of business server to the OS and communications app. The AI component to this deployment model is that recommendation engines provide business guidelines and data on previous transactions to help employees make informed business decisions. The device and communication platform flexibility are a benefit of this approach.
For now, the mobile device implementations of B2B chatbots, virtual assistants, and microservice architecture services are performing machine learning in the cloud. However, with smartphones including dedicated neural networking processors and supporting machine learning on-device, handsets from Apple, Huawei, and others are driving this intelligence to the network edge. Pattern recognition and usage analytics (that optimize the device memory while it is charging) are some of the initial on-device applications. However, mobile device AI implementations are far from perfect.
Enterprise organizations need to understand what side effects that the new AI functions introduce. For example, unlocking a smartphone with the user’s face sounds convenient, but the fact remains that it can be spoofed. Consider smart authentication policies, such as multi-factor authentication, rather than outright blocking face recognition.
The potential for AI in the mobile enterprise is truly exciting as it signals the beginning of augmenting worker intelligence. Yet many companies will struggle with AI technologies because they have not evolved to operating with ambiguous outcomes. Several AI-powered applications are currently available that benefit every industry and size of business, while reducing the risk and exposure to companies that are looking to build their AI competence.
Is your company ready for AI?