4 key job roles for success with artificial intelligence initiatives

The field of artificial intelligence has witnessed a huge renaissance in the past decade because of significant advancements in specialized computing, and better ways to collect, process and store large volumes of data. The continued diversity in AI projects, products and deployment models, combined with a requirement for rapid time to production, will require proper skills to meet these demands.

A recent Gartner survey revealed that leading organizations expect to double the number of AI projects in place within the next year. At this rate, the lack of AI skills will continue to be the No. 1 challenge for enterprises looking to succeed in their AI initiatives through 2025. Herein lies the value of engaging an AI architect and other related roles to do so.

Defining and Engaging the AI Architect
Today, enterprise architecture and technology innovation leaders, including CIOs and CTOs, view AI as a cornerstone of their innovation strategy. Although the 2019 Gartner CIO Survey reveals significant impending investments in AI and strong growth in deployments year-over-year, very few enterprises have realized its full potential because of process immaturity, diversity and complexity of the AI technology stack.

The presence of an architectural leader, or an AI architect, is an effective way to envision, build, deploy and operationalize an end-to-end machine learning (ML) and AI pipeline. AI architects play an important role by being the curators and owners of the architecture strategy. They are the glue between data scientists, data engineers, developers, operations and business unit leaders to govern and scale the AI initiatives.
cit developer.jpg
An AI architect is closely related to other architecture roles, such as enterprise architects and solution architects, but is laser-focused on the transformational architecture efforts that AI introduces. Responsibilities might include mapping current and future stakeholder requirements to technical implementation; developing prototypes; operationalizing AI.

Required skills for an AI architect
AI architects are expected to possess a wide array of technical and nontechnical skills that are often difficult to acquire over a short period of time. Figure 1 summarizes an ideal checklist of the key “hard and soft” skills needed to be successful in this role.
chart tech skills.png

Figure 1

Additional AI roles for successful AI and ML projects
Data science is a team sport, although many organizations start with a flawed assumption that data scientists alone are enough and can handle AI imperatives around data and analytics, machine learning operations and enterprise architecture. Apart from AI architects, the other important roles needed to succeed with AI projects are:

• Data engineers: They are responsible for making the appropriate data available for data scientists. Data engineers generally work on implementing complex projects with a focus on collecting, parsing, managing, analyzing and visualizing large sets of data.

• Data scientists: They are the backbone of AI and ML projects. Their responsibilities include choosing analytics challenges that offer the greatest opportunities, determining appropriate datasets and algorithms and identifying patterns and formulating solutions.

• ML (or AI) engineers: ML (or AI) engineers sit at the intersection of software engineering and data science. They are responsible for taking data science models and helping scale them out to production and ensure that business service level agreements are adhered to. They are also an important part of the continuous feedback loop that is vital for enhancing validity and accuracy of AI workloads.

Call to action for IT leaders
Gartner appreciates that the AI architect role will likely be a newly created role in most organization, and IT leaders will need to look internally or externally to fill it. IT leaders looking to groom or hire an AI architect should:

• Conduct a gap analysis of existing roles and skills within the organization and identify internal candidates who can be trained to this role—likely individuals who have an enterprise architecture, data science, infrastructure or data and analytics background.

• Ensure the selected individual engages in a variety of data science projects early on to gain practical experience and knowledge of the intricacies of AI.

• Stay abreast of new use cases, technologies, sourcing models, vendors and other market developments to succeed. AI is indeed a rapidly evolving field.