Skip to main content Skip to Footer

BLOG


Digital Perspectives
New views. Applied now.

Digital Perspectives

New Views. Applied Now.

June 07, 2019
Growing up: Getting to AI value at scale
By: Fernando Lucini

At CogX this week, hundreds of start-ups are gathering in King’s Cross to showcase some of the latest innovative and disruptive technology. Taking centre stage will be artificial intelligence (AI). And for good reason. By now, very few organisations are not exploring its potential. There’s no longer any question about the promise of AI. But, this year enterprises will be asking a very different question—how do we turn that promise into value at scale?

It’s clear from CogX that the challenge is not the technology itself. Achieving value at scale requires finding the right way to bring together the organization, talent and processes that can convert AI from tantalising prospect to real business growth.

Put value in the driving seat

First of all, it’s essential to define what value means for the organisation. With that clearly understood, AI can be targeted to address the most valuable opportunities and align to strategic priorities. Leading organisations make that their goal. They make sure that they focus on the most valuable AI interventions in business terms, driven by what they can deliver rather than what they can learn for its own sake.

Don’t reinvent the wheel

With the abundance of proven, low-cost AI solutions available in the market, it makes little sense to build solutions from scratch. Unless something is unique and so differentiated that it’s going to deliver multiples of value, the wiser option is to reuse, buy or partner. And buying does not have to mean investing in an enterprise software product. Using a service, open-source software or packages can all help achieve the goal of getting to value at speed.

Make the right connections

Companies will use different models to integrate AI in the best way to suit them. It could be that a centralised structure houses AI tools and teams. Or resources might be dispersed across business units around the world. Many businesses are finding that a combination of the two works best. This is a hub and spoke model that centralises AI governance and leadership in a centre of excellence, but connects to other parts of the business. That way, there’s consistent governance and standards supported by the centre, without inhibiting the development of AI across the wider business.

Evaluate quickly and decisively

A full pipeline of AI concepts and ideas that could create value is essential. But those ideas have to flow quickly and be rapidly, but robustly, evaluated. That requires a disciplined, iterative approach in place to shaping, developing and moving forward or abandoning an idea. That’s the first essential phase of an AI lifecycle, with the ultimate goal of being able to embed AI into every new product and customer offering.

A comprehensive AI operating model

A strong and scalable AI operating model spanning the entire organisation and way of doing business is also critical to scale AI value. This needs to take in, for example, how products and services will need to change, whether compliance and regulatory concerns need to be addressed as well as making sure data foundations are solid and how marketing operates in a world of real-time and automated customer targeting. As well as looking inside, the operating model needs to take in the wider ecosystem of partners and vendors that could help to drive value.

The right talent mix

To create AI products and services that can deliver value, the right talent is a must. But while they’re a sought-after and precious resource, data scientists on their own are not enough. Data and software engineers are critical, too. And they need to be available in the right numbers to get concepts into production at scale. And scaling requires non-technical inputs as well. That means getting designers and governance experts involved early on.

Build and sustain trust

AI will only flourish if it’s responsible. Bias has to be designed out. Teams must understand—and act on—the ethical implications of what they’re doing. And AI must reflect and uphold the organisation’s core values with honesty, transparency and fairness. Without those and the trust they build, AI won’t be able to scale. And it won’t deliver the huge value it promises.

Popular Tags

    More blogs on this topic

      Archive