One of the most challenging aspects of AI deployments has been the recruitment and retention of data science talent. Creating an AI-ready enterprise will require organizations to build a team of data custodians: experts at blending information sources, providing feedback on new data sources and doing the requisite analytics to address business problem issues that arise. To begin, an analytics team will consist of data scientists, AI engineers, data engineers, data governance experts and specialists in data entry. Teams can grow and evolve as the needs for AI expand
Much of the construction and optimization of AI algorithms s remains something of an art requiring experts to define use cases and think through the clinical workflows and how the algorithms can be incorporated into the daily workflows of clinicians. Demand for these skills far outstrips supply at present; according to some estimates, fewer than 10,000 people have the skills necessary to tackle serious AI problems. and competition for them is fierce among the tech giants. What’s worse is that not too many jobs in healthcare are being filled with AI skills in mind. This is both an indication of the slower adoption of AI in Healthcare but also that health systems are not thinking proactively about this when recruiting new talent. Figure 1 shows that healthcare is at the bottom of industries in terms of the advertised jobs requiring AI skills.
Managing AI requires new expertise and rigor. There is significant expertise needed designing, implementing and sustaining benefits. The tools and science are far more complex than logical decision support rules. The health system needed a central expert team and tools. An example is in Ohio Health, where the health system built a hub-and-spoke model where there was a central data scientist team, clinical informatics, program management and ongoing monitoring. The various business units and clinical project teams led the business case, workflow design and change management.
