IT Leaders: 3 Ways to Structure Successful AI Operating Models

By Anthony Mullen, VP Analyst at Gartner.

  • 10 months ago Posted in

 AI innovations continue to deliver big benefits to business and adoption rates will continue to accelerate in coming years.

However, AI technology maturation and the diverse approaches of AI make it difficult to capture and sustain value from AI initiatives.

By setting up effective AI operating models. IT leaders can realise successful AI initiatives.

Gartner identified three key ways for IT leaders to structure such operating models.

1. Benchmark Your Organisation’s AI Maturity Level

When it comes to structuring AI operating models, benchmark your organisations current maturity levels to maximise value enablement from your organisation’s current state.

Begin by benchmarking current maturity levels against internal requirements such as skills, capabilities and technologies and external factors such as maturity of industry peers, emerging AI innovations and disruptions and legislative activity.

Then partner with your senior leadership to assess the gap between organisational AI skills and the AI project pipeline. Determine the role of each core business function in driving AI initiative success by defining both areas of ownership and each role’s level of involvement.

Upskilling current employees through customised programmes by role is essential. Complement these reskilling efforts by creating environments that attract new AI talent to your organisation. Smart resourcing of AI projects requires utilising gap analysis – mapping project schedules to the availability of internal skilled resources. Prioritise the development of AI operationalisation skills to speed realisation of value from AI.

Focus on creating operation AI systems that manage multiple data, model and deployment pipelines to standardise data engineering, model engineering and deployment practices.

2. Create an AI lab to Identify AI Use Cases

Create an AI lab to identify AI use cases for your organisation and deliver a pipeline of AI pilot candidates to structure AI operating models. The AI lab should focus on agility and include an interdisciplinary team that nurtures implementation focused collaborations of diverse enterprise stakeholders. Teams need to rapidly explore proofs of concepts (POC), delivering into production successful AI pilots that are aligned with business value.

Accelerating successful productisation of AI pilots by establishing synchronous change management processes led by senior leadership is essential. To deploy AI, the CIO and relevant enterprise architecture teams need to be empowered to utilise modernised advance processes across data operations, cloud adoption and utilisation of microservices. If internal resources are constrained or unable to deliver initiatives in a timely manner, strategically outsource AI projects.

Create an AI management and governance function that owns responsibility for managerial and tactical execution of AI. The mandate of this function must include stakeholder collaborations, portfolio/project/programme management and flexible delivery modes.

3. Expand Use of AI Within the Organisation

Expand the use of AI within the organisation by supporting AI initiatives, encourage cross-functional collaboration, and embrace methods, techniques and processes that accelerate new projects. This will encourage cross functional collaboration, support embedded AI initiatives to develop assets, and ensure that AI is surfaced in line-of-work applications and interfaces.

Additionally, formalise a partnership with software engineering leaders who can get more models into production and develop a comprehensive process for scaling business-value-linked AI projects into production. This is done through iterative standardisation and automation of best practices across all AI projects.

Next, assign ownership of the outcomes to ensure long-term support, maintenance and manageability of AI products and utilise AI design patterns. Design patterns enable a common language for communication around the business and act as a starting point and accelerator for new projects.

A successful AI operating model embraces the core enterprise business strategy and coordinates multiple functions — D&A, enterprise apps and infrastructure — to capture, enable and sustain value from AI. Gartner analysts will explore how leaders need to structure their AI operating models at the Gartner Data & Analytics Summit, taking place in London, on 22 – 24 May 2023.

By Steve Young, UK SVP and MD, Dell Technologies.
By Richard Chart, Chief Scientist and Co-Founder, ScienceLogic.
By Óscar Mazón, Senior Product Manager Process Automation at Ricoh Europe.
By Chris Coward, Director of Project Management, BCS.
By Trevor Schulze, Chief Information Officer at Alteryx.