Accelerating and De-risking Artificial Intelligence and Machine Learning Maturity

This article is adapted from my session at the Enterprise Data & Analytics Online 2020 and shared within MinterEllison

At the recent Enterprise Data & Analytics Online 2020, our Head of Data and Analytics, from the Change and Transformation team, Shaheen Saud, shared the Firm's approach to Artificial Intelligence (AI) and Machine Learning (ML) experimentation as part of its Data and Analytics program and discussed some of the strategies organisations can employ to accelerate and de-risk the drive to AI maturity.

In sharing the MinterEllison story, Shaheen outlined his journey in joining the purpose-led, circa 200 year old, global professional services organisation 2 years ago. He described the Firm's 2025 strategy, which laid out a really compelling vision to modernise and digitise the Firm's business model and Data and Analytics as very much being a key enabler of that transformation.

As AI becomes more mainstream, and as organisations look to increasingly leverage AI to make better predictions and informed decisions, enabling innovation and productivity gains, Shaheen laid out three strategies to accelerate and de-risk the drive to AI maturity within our Data Analytics program:

1.    Outcomes focus

Having a focus on outcomes and what problem are we trying to solve helps AI-initiatives deliver tangible value. When the Firm thinks about goals and outcomes, we think about impacts on our Clients, our People and the Firm. For our Clients, the use of innovative solutions like AI is around offering market leading services. For our People, it's about greater job satisfaction and engagement and for the Firm, it is about delivering on business outcomes.

2.    Having a pragmatic roadmap

Shaheen highlighted the need for a pragmatic roadmap to help organisations mature their capabilities.

Whilst the 'horizons' or phases for a roadmap will be dependent on a particular organisation and their current levels of maturity and capabilities as well as aspirations, some of the key fundamentals remain consistent and include understanding and building trust in an organisation's data.

The roadmap should also help inform decisions around technology investment, including Cloud adoption and at what point an organisation pivots to AI experimentation.

The roadmap needs to be underpinned by a continuous focus on capability, talent, business sponsorship as well as return on investment.

3.    De-risked approach with high standards to build trust

AI experimentation often comes with a high degree of cost, risk and complexity. For organisations starting to explore AI, having a phased, business use case led approach and starting out with small teams built around 'data citizens' leveraging auto AI/ML platforms will minimise risk and accelerate adoption. An acceptance of the process of experimentation and un-perfect solutions it produces first time in order to learn and build on is key to this approach.

In addition, leveraging trusted technology partners to explore the art of the possible as well as identify business focused use cases with tangible benefits will help drive collaboration.

Furthermore, having really high standards, focusing on accuracy, governance and privacy, and making it easy to explain what's happening in the models builds trust in AI initiatives.