This paper presents the design and the CFD self-propulsion analysis of a ducted propeller to operate in a DARPA SUBOFF hull Autonomous Underwater Vehicle (AUV). The ducted propeller is of the pump jet type with 9 blades at the rotor and 11 blades at the stator. The interactive process of design and optimization uses the potential lifting line theory and CFD RANS analysis for obtaining the self-propulsion point, with the propeller placed behind the AUV hull. During the lifting line design, the rotor diameter, hub diameter, design rotation, blade section chord and length of the duct are modeled by a Kriging Metamodel technique and optimized through random sampling in order to maximize the quasi propulsive coefficient. The optimized configuration from the lifting line and Kriging Metamodel is analyzed using Ansys Fluent 2019 solver. The CFD analysis behind the hull allows including wake effects, thrust deduction factor and viscous effects directly into the model. The lifting line and CFD processes are used interactively to optimize the pitch, the circulation and the camber until the required thrust is achieved.