Hi, we are looking for someone who can help us create this. A mobile-optimized upper-body humanoid .glb model built specifically for motion retargeting rather than cinematic detail. The model should be authored in a neutral A-pose or relaxed T-pose and include a clean skeleton covering the full upper-body signing chain: chest, neck, head, shoulders, upper arms, lower arms, wrists, and complete finger chains for both hands, along with basic facial blendshapes for eyes and mouth. Because the target runtime is Expo/Three.js on mobile, the asset should remain lightweight: ideally one skinned mesh, restrained polygon count, limited materials, modest texture sizes, and no cloth, hair, or physics-driven secondary rigs. A predictable rig is more important than high visual detail. In practice, the most important qualities are clear bone naming, stable local bone axes, and a consistent thumb/finger hierarchy. For context: This is for Deaf education in India. This avatar would collect sign language from Deaf users and imitate it.