Machine Learning Engineer, Graph Deep Learning - Credit
Core
Develop graph deep learning models using large-scale user relationships and financial behavior data to improve credit risk assessment.
Role type
Machine Learning Engineer (Graph Deep Learning)
Builds
Graph deep learning models for credit risk, behavioral scoring, and application scoring
Domain
Fintech / Credit Risk / Graph Neural Networks
Deliverable
production ML models
Required skills
Graph Neural Networks (GNN), Graph Attention Networks (GAT), GraphSAGE, GCN, Python, PyTorch, DGL, PyTorch Geometric, SQL, experimental design, ablation studies, point-in-time data handling
Preferred skills
Credit risk modeling, heterogeneous graphs, temporal graphs, graph self-supervised learning, distributed training (Spark/Hive), large-scale graph processing
Technologies
PyTorch, DGL, PyTorch Geometric, Spark, Hive
Responsibilities
Design nodes, edges, and features from contact networks and transactions; Implement and improve graph attention models; Compare training strategies (pooled, continual, rolling); Incorporate graph scores into risk models; Collaborate on data processing and inference pipelines; Explore heterogeneous and temporal graphs for credit risk
Seniority
Mid-Senior, hands-on IC
