With the development of science and technology and the increasing complexity of macro decision-making, giving full play to new-generation information technology as represented by big data and artificial intelligence in the decision support system of think tanks, and realizing the organic integration of think tank experts’ knowledge and machine intelligence are the inherent requirements for the scientific and standardized decision-making consultation of think tanks. With the heavy involvement of think tank experts, the human-machine decision support system will organically integrate computer technology, AI technology, software engineering technology and modern soft science theories and methodologies e.g., management science, system science and policy science, to provide support for solving complex think tank decision-making problems. Therefore, this chapter, in the framework of the Double Helix Methodology in Think Tanks, will study the methodology and architecture of human-machine system for decision-making consultation in think tanks in a bid to provide theoretical and practical reference for the construction of decision-making system in think tanks.