算法工程专家（模型推理平台-国际支付）-Global Payment
Core
Design, develop, and optimize a unified model inference platform for international payment scenarios, supporting high availability and low latency for large and small models across multiple frameworks.
Role type
Senior IC model inference platform engineer
Builds
Unified model inference platform serving risk control and compliance systems in international payments
Domain
Fintech / International Payments / Distributed Systems
Deliverable
production ML models
Required skills
Model deployment and serving, Inference optimization (quantization, pruning, operator fusion), Multi-framework support (PyTorch, TensorFlow, ONNX, TensorRT), Distributed inference, High-performance backend development, System architecture design
Preferred skills
LLM inference platform experience, Payment/finance domain knowledge, Microservices architecture, Open source contributions
Technologies
PyTorch, TensorFlow, ONNX, TensorRT, vLLM, TGI, DeepSpeed, ONNX Runtime, gRPC, HTTP API, CI/CD
Responsibilities
Design core modules for model deployment and service orchestration, Implement inference scheduling mechanisms (load balancing, auto-scaling), Optimize model performance and cost efficiency, Build standardized service systems with monitoring and SLA guarantees, Collaborate on model deployment norms and automation pipelines
Seniority
Senior, hands-on IC