Remote PartTime Data Scientist Multimodal Foundation Model Evaluation and Data
Core
Design and execute evaluation pipelines, develop benchmarking methodologies, and create high-quality training datasets for multimodal foundation models.
Role type
Data Scientist (Multimodal Foundation Model Evaluation)
Builds
Evaluation frameworks, data pipelines, and training datasets for multimodal AI systems.
Domain
Artificial Intelligence / Multimodal Foundation Models
Deliverable
production ML models
Required skills
Machine learning theory, computer vision, deep learning architectures, statistical analysis, Python programming, PyTorch, TensorFlow, JAX, experiment design, data pipeline development
Preferred skills
PhD in quantitative discipline, experience with SAM/LLaVA/CGPT4V, statistical experiment design, cloud-based ML infrastructure, peer-reviewed research publications
Technologies
Python, Jupyter, Pandas, NumPy, Matplotlib, PyTorch, TensorFlow, JAX, AWS, GCP, Azure, Docker, Kubernetes
Responsibilities
Design and maintain evaluation frameworks for large-scale multimodal models; build data ingestion and transformation pipelines; conduct statistical analysis of model performance and bias; plan and execute systematic experiments including ablation studies; translate data-driven findings into feature specifications; create reusable software utilities for diagnostics and reporting.
Seniority
Mid-Senior, hands-on IC