Senior ML Engineer: Real-Time Personalization & Recs

ESPN; New York (NY); Sur place; USD 151 000 - 202 000; Plein temps

Résumé du poste

Disney Entertainment and ESPN Product & Technology is seeking a Senior Machine Learning Engineer in New York, NY. You will design, develop, and productionize deep learning models for large-scale ESPN recommender systems.

You will own model architecture, training methodologies, offline evaluation, and online serving, while collaborating with platform and data teams to ensure reliable performance at scale.

Qualifications

• Strong knowledge of deep learning theory and architectures.
• Experience designing and launching production-ready deep learning models.
• Hands-on experience with recommendation/ranking models at scale.
• Proficiency in PyTorch or TensorFlow and Python software engineering.
• Experience with large-scale training, distributed data processing, and evaluation.

Responsabilités

• Design and implement deep learning models for retrieval, ranking, and personalization.
• Translate product goals into model formulations and evaluation plans.
• Develop production-grade training and inference code with robust testing and reproducibility.
• Optimize models for latency, throughput, and memory usage in online serving.
• Own model productionization path from prototype to live deployment.
• Collaborate with data science, product, and platform teams on deployment.

Connaissances

Deep learning

Production ML

Python

PyTorch/TensorFlow

Inference optimization

Distributed training

Experimentation

Communication

Formation

Bachelor’s or Master in CS/ML

Outils

ONNX

TensorRT

CUDA

Custom kernels

Description du poste

Disney Entertainment and ESPN Product & Technology is seeking a Senior Machine Learning Engineer in New York, NY. You will design, develop, and productionize deep learning models for large-scale ESPN recommender systems.

You will own model architecture, training methodologies, offline evaluation, and online serving, while collaborating with platform and data teams to ensure reliable performance at scale.