In an exclusive interview conducted for ai:sight, Berryman shared his insights on the successful deployment of artificial intelligence (AI) at scale and the sustained value it brings over time. How do you balance short-term wins and long-term strategic goals when building an analytics roadmap? Y ou need to have balance in your pipeline. W hat I mean by that is sometimes you need to deliver short - term wins to earn the right to go after long - term strategic goals. Y ou want to enable rapid experimentation, demonstrate your capability and brand, and create attractive financial returns in an area important to the organi z ation. W hat ' s important to reali z e is that even if you have long - term initiatives in your pipeline, it ' s still essential to brea k them down into short - term milestones to demonstrate measurable progress. In D ecision Analytics, we brea k down our initiatives into two - or three - wee k sprints. How can businesses maintain an agile and adaptable analytics j ourney that responds effectively to changing mar k et dynamics and evolving business needs? the long - term ambitions of the AI pro j ect. W e have a separate 'S wat ' team designed to go after these urgent, short - term re q uests and enable rapid experimentation. T his, teamed with our close collaboration with multiple partners, gives us the scalability and fl exibility to meet demands rapidly. It ' s also important to recogni z e that urgent re q uests will always come up – these need to be balanced carefully with What measures do you implement to ensure the ongoing optimization and maintenance of AI systems to prevent value erosion and maximize long-term value realization? strategy of a particular business. T hey ' re also responsible for translating what we can deliver into a language the business understands. T here is one common thread that features in every pro j ect we wor k on, though – and that ' s the need to connect to business strategy and business outcomes first, which drives the initiatives we focus on. W e have engagement leaders aligned by business & functional area who identify how advanced analytics can help accelerate the vision and T he businesses we support have varied degrees of maturity in many different industries. S ome are much more sophisticated than others ; some employ data scientists, others don ' t. T his means we need an agile operating model to adapt to that – because there is no one - si z e - fits - all approach . O ne of the misconceptions is that most value comes from the minimum viable product ( M VP ) development. S o, you ' ve got a problem, and then build an AI solution to solve it.
How do you address the ethical considerations and potential biases when deploying AI at scale to prevent value erosion and maintain fairness? T o prevent value erosion, we need the ability to monitor what ' s put into production. S o, after production, there ' s a crucial step: monitoring the outcomes. W e have a separate team that monitors business outcomes while k eeping a close eye on data q uality and the performance of models. If we are not achieving business outcomes, then there ' s a problem. W e need to pic k up business process changes before they occur and identify when something isn ' t wor k ing. But this is the easier part. In my opinion, the last mile generates most of the value – the adoption phase — integrating the insights into the business process and scaling and sustaining this over time. It ' s also important to involve end users in the AI development process from the beginning to the ideation and scoping stages. If you don ' t do this, the danger is that you build a solution that no one wants or uses, wasting time and money. There's a lot of discussion around generative AI today. What should businesses consider when deciding whether to deploy these solutions? essential to have humans involved so that they can highlight these sorts of issues. W e also follow a tried - and - tested framewor k developed by the Institute of E thical AI and M achine L earning. T hese are important considerations. S ay you wanted to create a model to detect whether a student enrolled in a data science program will get an A. If the class consists of 90% boys, then the model is li k ely to predict that males are more li k ely to be successful. T his is because of bias in the population sample. S o, it ' s important to remove the demographic that is causing the bias. T his is why it ' s G enerative AI seems to have attracted more attention than anything else I ' ve seen in the realm of AI, so we are using that as a catalyst to facilitate conversations around more general AI. Because what most people don ' t reali z e is that generative AI seldom operates in isolation. D on ' t be surprised if it only ma k es up 10% of a pro j ect ; the other 90% is traditional AI. Businesses also need to consider the privacy and security of their data when employees start using generative AI solutions. F or example, third party managed platforms can provide a secure tenant in a secure environment. T his connection means that prompts entered into the solution are not shared for the training of the model – which means we can use it securely. T hat said, if you employ 3 0 , 000 people, you can ' t stop them from entering information into a generative AI solution from their phone, for example. S o, the best thing we can do in this situation is to educate people about the dangers of entering private data in a public environment.
3 8
Made with FlippingBook - PDF hosting