About

We are looking for a skilled 3D technical artist with experience in mesh processing and scan alignment. The task is to take an existing high-quality human body mesh template and produce a canonical template mesh that: Maintains identical topology (exact vertex count, order, and connectivity). Is aligned to a reference dataset of full-body human scans in a standard A-pose. Is exported as a clean, watertight OBJ file for downstream use in machine learning and simulation pipelines. This is not a creative character modelling project — the focus is strictly on technical precision in mesh handling. Responsibilities Import the provided base body mesh template. Use appropriate tools (e.g., R3DS Wrap, Blender, Maya, non-rigid ICP workflows) to: Align the template mesh to match the reference pose used in our scan dataset. Ensure the topology is unchanged throughout (same vertex count and ordering). Deliver a clean canonical OBJ file suitable for use as the foundation in automated pipelines. Document the workflow so it can be repeated consistently in the future. Requirements Proven experience with mesh wrapping, retargeting, or scan-to-template workflows. Deep understanding of 3D topology and why vertex order/consistency matters. Proficiency in one or more of the following tools: R3DS Wrap (preferred) Blender / Maya with shrinkwrap / non-rigid alignment ZBrush (for verification, not retopo) Python-based ICP or mesh alignment pipelines Strong attention to detail and ability to follow strict technical specifications. Ability to deliver quickly and communicate clearly. Deliverables A single watertight OBJ template mesh with: Identical vertex order and topology to the source mesh. Pose aligned to the reference dataset (A-pose). Clear written notes or a short video walkthrough of the workflow used. Bonus Skills (not required) Experience working with large human scan datasets. Familiarity with ML/AI pipelines that consume 3D meshes. Background in simulation, VFX, or 3D scanning. Project Scope One-off alignment task to start (single canonical mesh). Potential follow-on work adapting the same workflow to larger datasets. NDA agreement required.