Laurent Fresnel, CIO, The Star Entertainment Group
Artificial Intelligence (AI) deployed in the right area of a business can yield not only cost savings but, more importantly, a competitive advantage. The insights AI can reveal are superior to what any human is capable of achieving alone. The key is to deploy AI technologies within the enterprise only after you’ve identified and targeted specific use cases and the measurable outcomes you want to achieve. At The Star, we have identified specific use cases where AI will help augment human decision-making to improve the customer experience, reduce costs, and increase revenue. However, the potential business benefits must not overshadow the complexity of the technologies to be deployed and a clear understanding of the risks in using an augmented-AI solution.
In a pre-COVID world, most of the use cases across a broad range of industries were based on the premise of staff reduction. AI-enabled robots in manufacturing replacing blue collar jobs have been the most visible implementations. In the post-pandemic world where millions of jobs worldwide are being lost, the need for augmented AI in all part of an enterprise will evolve from a‘nice to have’ to a ‘must have’. Companies becoming more digital with fewer people will have no choice but to invest in augmented AI.
Our journey with AI started a few years ago with a single premise. As a large-scale integrated resort, the need to create one-to-one personalised customer experiences in real-time would require a technology ecosystem powered by an AI brain. We embarked on a journey to modernise our technologiesand edge closer to this goal. We went on to digitize all our products and customer interaction points. We still have a long way to go but for some of those products we deployed an AI solution using computer vision at scale.
This was unique, a world first in the sector, and has provided insights allowing us to make profound changes to our business. We also built an event-driven architecture that could connect the data capture by those digital products in real-time to our backend platforms and more importantly, in the future, our AI brain. Along this journey, we were able to create new products for our business which benefited from having this data flow in real-time. Accessing the data in real-time has enabled us to identify persons who have just arrived at our integrated resorts and, based on their loyalty tier status, customize their journey of interactions from the valet to the concierge to the hotel check-in desk. We have built a new platform which is collecting all data points from all digital interactions and products. This is a platform that has not only been able to provide valuable operational data to our digital products but also new advanced analytics capabilities to refine all parts of our business from workforce management to customer insights. The next step of our journey is to use AI models to recommend the next actions to our front-line people to better serve our customers.
This is a critical phase to deliver on the goal of creating personalized experiences at scale through the understanding of every single customer interaction. This is the time where AI and humans work together in a seamless manner. To get to this point, the human will have to trust the AI. We are planning to build the trust by embedding a ‘recommendations’ feature into every front-line digital touch point and to also provide some transparency on why the AI is recommending this interaction. We believe this will help our staff trust what the AI is recommending rather than relying on their own gut feeling. The premise is that our AI will be able to digest a lot of information in real-time based on a customer’s past interactions and, up to the last minute, be able to recommend a set of actions and options that optimize customer and business value. We also plan to track when a staff member will choose the AI-recommended action versus their own so we can further understand who was correct. We can then either convince our staff to trust the AI or improve our AI models.
In the post-pandemic world where millions of jobs worldwide are being lost, the need for augmented AI in all part of an enterprise will evolve from a‘nice to have’ to a ‘must have’.
