Artificial Intelligence (AI) strategy: 4 priorities for CIOs

Extracting the value of artificial intelligence requires gaining quick wins while developing at enterprise scale. Consider focusing on these key AI areas
98 readers like this.

It’s an exciting and scary time to be a technology leader: Exciting for the endless opportunities offered by rapidly evolving digital technologies – and scary due to the associated feeling of FOMO (fear of missing out).

Consider Artificial Intelligence (AI). Driven by the desire to tap unprecedented volumes of data for a broad array of real-world applications, many organizations see AI as a magic wand that CIOs can swing to generate customer delight and executive exhilaration.

CIOs know better, of course. The challenges that come with any new technology hit technologists harder and faster than the optimism driving it. This is especially true with AI and related areas such as machine learning (ML), data science, deep learning, natural language processing (NLP), and cognitive intelligence. Not only is talent scarce in these fields, but their vocabulary and application development are also different.

[ Check out our primer on 10 key artificial intelligence terms for IT and business leaders: Cheat sheet: AI glossary. ]

Extracting value from AI is about making a real-world impact with demonstrable quick wins while developing the organization for an enterprise-wide scale. Let’s look at the four key focus areas that can help make that happen.

1. Deliver differentiated offerings with AI-based, real-time decisions

One of the most critical priorities is identifying high-impact areas with opportunities to embed AI-based, real-time decisions in business processes. The ability to process contextual information in real time to make on-the-fly decisions is a powerful way to differentiate products, services, and experiences in the crowded marketplace.

For example, insurance firms can automate claims processing for real-time approvals based on pictures and videos provided by the claimant right from the place and time of the incident. Lenders can analyze risks in real time based on collateral and background information to offer on-the-spot loan approvals. Organizations can personalize and customize products and services across a broad array of use cases through the judicious injection of AI in their business processes.

The key lies in identifying several immediate-impact areas and focusing on creating a visible and measurable impact in customer experience – compared to, for example, plugging in chatbots just because everyone else is doing it.

[ Want to accelerate application development while reducing cost and complexity? Get the eBook: Modernize your IT with managed cloud services. ]

2. Implement AI engineering/MLOps to operationalize AI at enterprise scale

Gartner research shows that only 53 percent of projects make it from artificial intelligence (AI) prototypes to production. CIOs and IT leaders find it hard to scale AI projects because they lack the tools to create and manage a production-grade AI pipeline. This is a critical bottleneck because business processes cannot leverage AI capabilities effectively unless the engineering processes are mature enough to create a consistent pipeline of deployable models, regardless of investment, research, and proofs-of-concept.

Since AI engineering differs from “traditional” software engineering, CIOs must establish a strategy to institutionalize AI and ML methodologies. Many enterprises have found that the most effective way to do this is to establish a robust platform supported by a governance model.

A platform – let’s call it a unified platform because it combines various aspects from experimenting and design all the way to deployment – is a powerful mechanism. It enables CIOs to focus on engineering aspects of AI in a centralized manner, supported by a roadmap. It facilitates gradual scaling up without losing direction, while implementing business use cases and securing quick wins.

3. Leverage a cloud-based AI platform for flexibility and scalability

According to McKinsey’s survey, The State of AI in 2021, high-performance organizations in AI use cloud infrastructure much more than their peers do: 64 percent of their AI workloads run on public or hybrid cloud, compared with 44 percent at other companies. This high-performance group is also accessing a wider range of AI capabilities on the cloud compared to their peers. This is a critical factor because up-front infrastructural investment is one of the most significant deterrents for AI progress in enterprises.

[ Read also: 5 things CIOs should know about cloud service providers. ]

A cloud-based AI platform provides the flexibility to start small and experiment by focusing resources on building models and earning quick wins.

A cloud-based AI platform provides the flexibility to start small and experiment in a demand-driven manner by focusing resources on building models and earning quick wins, then scale upon realization of value. A cloud-based platform also enables organizations to focus on business value by abstracting out all the technological and engineering aspects within the platform. This “experiment, pilot, and scale” strategy goes a long way in navigating the tricky early landscape in the AI journey.

4. Prepare for enterprise-wide scaling of AI by enabling citizen modeling

Aside from infrastructure, another deterrent may be the expertise-oriented data science and modeling domain. The vocabulary and tools may be inaccessible to people outside the core expert group. Interoperability and deployable models may also limited due to the generally unfamiliar lexicon.

In order to fully harness the potential of AI at an enterprise scale, the platform needs to be accessible to business users, domain experts, and citizen developers from other areas, so they can collaborate in developing value-driven AI assets.

AI is a long-term game

These initial steps go a long way in establishing the long-term capability of an AI organization. Keep in mind that AI and related technologies are still evolving rapidly. Enterprises that keep pace will be best positioned to win in the long term. A platform and governance-based approach is the way forward.

[ What should you know about  AI/ML workloads and the cloud? Get the eBook: Top considerations for building a production-ready AI/ML environment. ]

Anurag Shah is Vice President – Head of Products and Solutions, Americas, Newgen Software. He also leads GSI relations as well as the consulting and pre-sales in the Americas. He has been with Newgen for over 22 years. In his previous role, he led and managed delivery and professional services for enterprise customers.