Machine Learning Engineer

We’re looking for a Machine Learning Engineer with a strong background in Computer Vision and Generative AI to join our R&D team. You’ll build and optimize pipelines for virtual try-on, pose-guided image generation, and garment transfer systems using cutting-edge diffusion and vision models.

Must-Have Skills

Core ML & Engineering

• Proficiency in Python and PyTorch (or JAX, but PyTorch preferred)
• Strong understanding of CUDA and GPU optimization
• Ability to build exportable, production-ready pipelines (TorchScript, ONNX)
• Experience deploying REST inference services, managing batching, VRAM, and timeouts

Computer Vision

• Hands-on experience with image preprocessing, keypoint detection, segmentation, optical flow, and depth/normal estimation
• Experience with human parsing & pose estimation using frameworks such as HRNet, SegFormer, Mask2Former, MMPose, or OpenPifPaf
• Bonus: familiarity with DensePose or UV-space mapping

Generative Models

• Strong practical experience with diffusion models (e.g., Stable Diffusion, SDXL, Flux, ControlNet, IP-Adapter)
• Skilled in inpainting, conditioning on pose, segmentation, or depth maps
• Understanding of prompt engineering, negative prompts, and fine-tuning for control

Garment Transfer Pipelines

• Ability to align source garments to target bodies via pose-guided warping (TPS/thin-plate, flow-based) or DensePose mapping
• Must ensure preservation of body, skin, hair, and facial integrity

Data & Experimentation

• Experience in dataset creation and curation, augmentation, and experiment reproducibility
• Competence in using W&B or MLflow for experiment tracking and DVC for data versioning

Nice-to-Have

• Understanding of SMPL rigging/retargeting and cloth simulation (blendshapes, drape heuristics)
• Experience fine-tuning diffusion models via LoRA or Textual Inversion for brand or style consistency
• Familiarity with NeRF or Gaussian Splatting (3D try-on and rendering)
• Experience with model optimization for mobile/edge deployment (TensorRT, xFormers, half-precision, 8-bit quantization)
• Awareness of privacy, consent, and face-handling best practices

Tools & Frameworks

• PyTorch, diffusers, xFormers
• OpenCV, MMDetection, MMSeg, MMPose, or Detectron2
• DensePose / SMPL toolchains
• Weights & Biases, MLflow, DVC

We Offer

• Opportunity to work on cutting-edge generative AI applications in computer vision
• R&D-focused environment with freedom to explore, test, and innovate
• Competitive compensation and flexible work structure
• Collaboration with a team of ML engineers, researchers, and designers pushing boundaries in human-centered AI