The road to No-Human-in-the-Loop Design is lined with opportunities but also pitfalls. Machine-learning (ML) is emerging as a powerful technology with applications in many high-complexity, high-cost chip design areas. This new, digital intelligence will make it possible to develop more predictive engines and enable tools to become self-learning. Yet, to fully realize the potential of ML, tomorrow’s design tools will need to expand well beyond today’s algorithmic approaches and enable data-driven decisions by managing multiple streams of data. For example, ML can guide designers in choosing the right timing constraints and tool settings by analyzing previous design data, netlists, and technology libraries. Models will be able to guide design flows to change the course of optimization and achieve optimal QoR for timing, power and area. For this to happen, design organizations will need to embrace data-driven practices that are currently being adopted in industries like IT, wholesale/retail, and transportation.