Principal ML Engineer

About the Role

Lead the development of  multimodal RAG system that transforms fashion sketches into parametric sewing patterns. This role combines cutting-edge computer vision, NLP, and domain-specific code generation to automate pattern-making for the fashion industry.

Key Responsibilities

Technical Leadership

• Architect and implement end-to-end multimodal RAG pipeline for sketch-to-pattern conversion
• Design and optimize computer vision models for fashion sketch analysis and feature extraction
• Lead development of custom LLM fine-tuning for Ruby algorithm generation on AWS infrastructure
• Establish MLOps best practices using AWS AI services for model deployment, monitoring, and continuous improvement
• Define technical strategy for scaling ML infrastructure leveraging AWS GPU clusters and cloud services
• Implement automated training and deployment pipelines for continuous Ruby code generation improvement

Research & Innovation

• Research state-of-the-art techniques in multimodal learning and cross-modal retrieval
• Develop novel approaches for fashion domain-specific pattern recognition
• Pioneer techniques for parametric algorithm generation from visual inputs

• Collaborate with fashion designers to understand domain-specific requirements and constraints

   

Team Leadership

• Mentor senior and junior ML engineers in advanced techniques
• Drive technical decision-making across the ML team
• Collaborate with product, engineering, and design teams to align technical capabilities with business goals
• Establish coding standards and review processes for ML codebase
• Lead hiring and onboarding of additional ML talent

Required Qualifications

Technical Expertise

• Extensive experience in Machine Learning, Computer Vision, or related field, or equivalent industry experience (8+ years)
• Deep expertise in multimodal learning, particularly vision-language models
• Proven experience with RAG systems, vector databases, and large-scale embedding retrieval
• Strong background in computer vision: CNNs, Vision Transformers, object detection, image segmentation
• Advanced knowledge of NLP and LLM fine-tuning techniques (LoRA, QLoRA, RLHF)
• Production experience with PyTorch, transformers library, and modern ML frameworks
• Experience with open-source LLM architectures and deployment at scale
• Expertise in designing and implementing custom training pipelines for domain-specific applications

Preferred Qualifications

• Previous experience in computer-aided design (CAD) or algorithmic content generation
• Understanding of parametric modeling concepts and constraint-based systems
• Interest in fashion technology and pattern-making processes
• Experience with code generation for domain-specific applications

What We Offer

• A fully remote work environment with flexible hours.
• A collaborative, innovative team with growth opportunities.
• Compensation for sick leave and regular paid vacations.
• Holidays according to the US calendar.