Relational Foundation Model Engineer, Modern Data Stack
Core
Design, build, and evaluate novel Transformer and graph neural network architectures for a unified foundation model that understands relational databases and heterogeneous graphs.
Role type
Senior IC machine-learning engineer (relational foundation models)
Builds
Production systems for recommendation, forecasting, fraud detection, and predictive maintenance
Domain
AI/ML, Graph Learning, Relational Databases
Deliverable
production ML models
Required skills
Transformer architecture design, Graph Neural Networks (GNN), Python, PyTorch, Large-scale model training, Post-training optimization, High-performance inference
Preferred skills
Graph-based machine learning, Publications at NeurIPS/ICLR/ICML
Technologies
PyTorch, Transformers, Graph Neural Networks
Responsibilities
Collaborate with researchers to enhance models for relational schemas and heterogeneous graphs; Tackle real-world challenges in forecasting and fraud detection; Contribute to scalable and adaptable ML solutions; Work across the full lifecycle from architecture design to inference acceleration.
Seniority
Senior, hands-on IC
