Sr. ML Infrastructure Engineer II, Personalization
Core
Design, train, and ship recommendation models (retrieval, ranking, re-ranking) and build production ML systems for Slickdeals' personalization stack.
Role type
Senior ML Infrastructure Engineer (Recommendation Systems)
Builds
Recommendation models and production ML pipelines serving tens of millions of users
Domain
E-commerce / Personalization
Deliverable
production ML models
Required skills
Recommendation system design, deep learning for recsys, model evaluation methodology, A/B testing, distributed computing, cloud data processing, ML modeling frameworks, model serving platforms, vector retrieval/ANN, Linux/Ansible/Docker/Kubernetes, AWS infrastructure, hardware/resource management
Preferred skills
Feature stores, real-time/streaming feature engineering, LLM-augmented retrieval, e-commerce recommendation domain
Technologies
PyTorch, TensorFlow, AWS SageMaker, vector databases, Elasticsearch, HBase, SQS, Kafka, REST, LLMs, TorchServe, TensorFlow Serving, NVIDIA Triton, FAISS, ScaNN, OpenSearch, Pinecone, Weaviate, Apache Spark, Presto, SQL
Responsibilities
Design and train two-tower/dual-encoder retrieval and neural ranking models; build embedding pipelines; improve candidate generation strategies; define and run offline/online evaluations; build end-to-end ML pipelines for data preparation, training, deployment, and monitoring; design low-latency model serving; build feature pipelines and feature stores; design reliability and observability infrastructure; improve training cost and reproducibility
Seniority
Senior, hands-on IC