Designing the AI Experience
Background & Solution
The executive team of an energy client set the strategic goal of doubling output in 5 years.
To achieve this objective, several things needed to happen starting with problem definition. A variety of methods performed inside the business units included workshops, interviews, and surveys critical to the strategic goal—which identified two breakthroughs:
The first breakthrough uncovered was that the barriers to achieve the strategic goal shared the following themes within all the business units:
Aging, obsolete, and inefficient work processes
Insufficient digital skills among the workforce
Poor data access across the enterprise, leading to lower quality and slower decision making
By distilling these challenges for the business units into shared fundamental problems, large scale changes could then be implemented to address a host of specific issues.
The second breakthrough was discovering that the critical work path of activities and decisions contributing to output were ripe for digitilization. This could be best achieved by leveraging an existing AI technology whose development was in flight, and then supporting it with digital reviews, hand offs, and co-production of deliverables that collectively cut the cycle time in the entire workflow and value chain by more than half.
By studying how people were working, we were able to define the future state, and in doing so identify and align all relevant parties. All that was left was to design the AI experience core to the transformation effort in a manner compatible with ongoing sprint cycles and agile development. More specifically, the nature of the AI program would be trained to recognize shapes and images, so that the reduction in human effort would come from the AI program searching enterprise data and returning congruent shapes and images that it was trained to find, eliminating the effort and time for a human to do the same work. The AI program would then leverage the benefits of data consolidation projects that were already ongoing.
With the client’s development teams in mind, Daito defined the AI experience by concepting and rapidly prototyping a variety of training and data retrieval experiences from low, to medium, and finally to high fidelity prototypes while collecting feedback from end users, SMEs, and product owners. The result was defining an easy-to-use experience for an aging workforce with outdated digital skills, while at the same time providing a consumer-grade design experience that was attractive to the young millennial workforce.
The best approach was to let people and business needs shape the needed technology. From a thorough user-centered design process, the needed future state for the entire value stream was clearly defined, designed, and planned in such a way that led to better results than a technology-first approach which would have suggested buying an AI solution and then finding a problem for it. As data shows, technology first approaches typically result in a waste of money and effort.
Beginning with a business and employee-first approach ensured we were able to identify the full scope of not only what was needed to optimize the workflow, but also all the necessary feature requirements to fully support the AI program. This process resulted in us focusing on high value specifics like remote handoffs, reviews and approvals, and co-production of critical deliverables that collectively contributed to a successful doubling of output.