Despite the promise of AI, 90% of data science initiatives fail to achieve adoption and hence value. Even with adoption, 70% of models don’t get updated and lose relevancy. At the end of the day, Gartner predicts that over 50% of CDOs will fail within their first year of the job. Why?
For organizations to build data products as capabilities that drive value, we need to lead with mindfulness and human centric design — begin with the end in mind with the human in the center. Engineering and data science on its own will not drive success because we are in the business of human transformation. ML Ops should have (human-centered design) HCD infused E2E: from strategy and problem definition and to the intended, implemented experiences.