Insights: Building a value-added culture of everyday AI
Now Reading
Insights: Building a value-added culture of everyday AI

Insights: Building a value-added culture of everyday AI

Here are the six key factors that can help the process

Everyday AI

It is true that artificial intelligence (AI) will be the great enabler in the years to come and that regional organisations have come to believe this en masse. But we must ask ourselves how many of these deployments went according to plan. How many of them added value? Here, I share the six main drivers of success in becoming an AI-mature organisation with a culture of everyday AI.

01. Value
Through implementation of a use-case qualification framework, organisations should appoint people to project roles that can champion change that focuses on value creation. Domain experts should steer projects, as they are the closest to the issues that AI should seek to address. At this stage, KPIs must be agreed. KPIs should, where possible, be designed to quantify value beyond the use cases themselves, venturing into areas such as reuse and capitalisation across projects.

02. Vision
During every stage of every phase – design, implementation, and the rest – the organisation’s leaders should get together in regular executive workshops for knowledge transfer and ideation. Each use case that is implemented will be done in consultation with the map. Different projects will belong in different phases according to the organisation-wide capabilities that have been built up to that point.

03. Data
Enterprises should commit to the democratisation of data access. Non-IT employees have a lot of insight to offer if they are granted insights themselves. By understanding key metrics and gaining an overall view of the corporate strategy on AI, staff outside the core project teams could happen upon an idea that might otherwise be missed. Such access requires a robust infrastructure and clear metrics around the accuracy and useability of data.

04. Systems
If everyday AI is to work, then the infrastructure must support the tools and platforms that are being introduced. Infrastructure transformation must be designed in parallel with AI projects but implemented ahead of them. It should support the bandwidth and scaling requirements of everyday AI in terms of both compute and storage capabilities and be implemented in such a way that the organisation avoids technical debt down the road.

05. Talent
The region suffers from skills gaps in technology fields. While a range of government initiatives have been devised to solve this problem, many of these may take several years to bear fruit. In the here and now, enterprises that want to build an everyday AI culture can look to their own talent pool. The challenge is in driving the AI programme forward while delivering an employee experience that can attract and retain the right skills to make it happen. One way of doing this is to appoint citizen developers – non-technical staff who build AI solutions using low-code platforms. This gets the AI initiative up and running while creating a rewarding environment for non-IT staff and relieving AI specialists and data scientists of the workload.

06. Governance
Good AI governance ensures that data is useable, accessible and secure. It strikes a balance between auditability and permissions that allows the organisation to scale its use of data. If the progress of AI systems is not monitored, performance can degrade. People must be appointed to ensure responsible AI – to measure the progress of models and ensure their operation is not becoming detrimental to the business and other parties.

Gregory Herbert is the SVP and general manager – EMEA at Dataiku

Read: The AI talent factory – what it is and how to build it

You might also like


Scroll To Top