Salary
💰 $60 - $80 per hour
Tech Stack
CloudPythonPyTorchTensorflow
About the role
- <div><!--block--><strong>Position Type:</strong> Contract/Affiliate Resource Pool <strong>Duration:</strong> Immediate need: Remote, 20-40 hours over 6 weeks. Thereafter, project-based engagements (2-6 months typical) <strong>Location:</strong> Remote/Hybrid (travel to client sites as needed)<br><br>We are seeking a highly skilled Multimodal AI Training Data Scientist to join our elite resource pool of AI transformation specialists designing and implementing cutting-edge multimodal AI solutions. The ideal candidate will drive enterprise AI initiatives that solve complex technical document understanding problems and transform how organizations interpret critical engineering and industrial drawings.<br><br> </div>
Requirements
- Multimodal AI & ML Foundations. 5+ years architecting and deploying vision–language models (VLMs) for technical document analysis in production, with deep expertise in computer vision, NLP, and multimodal AI system design.\n
- Fine-Tuning & Model Optimization. Proven track record fine-tuning large language models (e.g., Google Gemini, GPT-4V, Claude) and implementing advanced computer vision techniques such as object detection, symbol recognition, and document analysis for complex technical drawings and diagrams.\n
- Training Data Engineering. Expert in designing instruction-following datasets, conversation formats, and robust quality assurance frameworks to ensure technical accuracy, aligned to business requirements.\n
- Document AI Solutions at Scale. Enterprise-level experience developing PDF processing pipelines with OCR, image analysis, annotation workflows, structured data extraction, and ML training infrastructure that deliver measurable business value.\n
- Programming & Deployment. Python mastery with ML libraries (PyTorch, TensorFlow, OpenCV, transformers), cloud platforms, and production-grade AI deployment using Agile methodologies.\n
- Collaboration & Communication. Ability to collaborate with subject matter experts to rapidly acquire domain knowledge in technical or regulated industries with strong written and verbal skills.\n
- Eligibility. Canadian or U.S. citizen, or visa holder authorized to work in the U.S. (no sponsorship available).