Senior AI/ML Engineer - Model Training & Optimization

About the role

LookRoom AI, a US-based AI startup building a personal assistant for digital wardrobes and personalized outfit suggestions, is looking for a Senior AI/ML Engineer to join its founding-stage technical team. The engineer trains, fine-tunes, evaluates and optimizes the AI models behind three core technologies: avatar creation, virtual try-on and personalized outfit recommendations. The position is equity-based and starts part-time and hybrid; it may become full-time and on-site later.

Key responsibilities

• Design and run training and fine-tuning workflows, select pretrained models and build reproducible experiments and pipelines.
• Perform hyperparameter tuning, benchmarking and model validation, and track experiments, datasets and model versions.
• Evaluate accuracy, visual quality, relevance, latency and computational cost, and optimize GPU utilization, memory and inference performance.
• Support model deployment, monitoring and continuous improvement, and recommend practical solutions to technical limitations.
• Collaborate with data scientists, 3D artists, backend developers, product and founders.

Required profile

• 5+ years of hands-on AI/ML engineering or applied machine learning experience.
• Strong Python, PyTorch and/or TensorFlow, and experience training and evaluating deep learning models.
• Experience with computer vision, GPU training and inference, and data preprocessing and augmentation.
• Ability to work with backend engineers on model serving and API integration.
• Good written and spoken English; comfort with early-stage startups and evolving requirements.

Required skills

• Neural networks, optimization algorithms and model architectures
• CUDA, mixed-precision training and performance optimization
• Git, Docker, Linux and cloud-based ML environments
• Highly valued: OpenCV, MediaPipe, 3D reconstruction, diffusion models, virtual try-on, recommender systems, MLOps tools such as MLflow or Weights & Biases, and ONNX/TensorRT

Applicants should send their CV, LinkedIn, GitHub or project portfolio and brief answers to questions about their trained models, computer vision and try-on experience, recommendation systems, GPU optimization and production evaluation.