Sr. ML Infrastructure Engineer II, Personalization

About Slickdeals:We believe shopping should feel like winning. That’s why 10 million people come to Slickdeals to swap tips, upvote the best finds, and share the thrill of a great deal. Together, our community has saved more than $10 billion over the past 26 years.We’re profitable, passionate, and in the middle of an exciting evolution—transforming from the internet’s most trusted deal forum into the go-to daily shopping destination. If you thrive in a fast-moving, creative environment where ideas turn into impact fast, you’ll fit right in.The Purpose:The Personalization team owns the systems that decide what each Slickdeals user sees, from homepage and feed rankings to deal recommendations across the site and in lifecycle channels. Personalization is one of our highest-leverage investments: it directly drives engagement, retention, and revenue across tens of millions of monthly users.We’re hiring a Sr. ML Engineer II who can operate end-to-end across the recommendation stack. This is a true hybrid role with roughly half modeling and half infrastructure. You will design and ship recommendation models (retrieval, ranking, and re-ranking) and build the production ML systems that train, serve, and evaluate them at scale. You’ll work closely with data scientists, product engineers, and the Search & Discovery and Shopping Graph teams.You will be building products using technologies such as AWS SageMaker, PyTorch, TensorFlow, vector databases, Elasticsearch, HBase, SQS/Kafka, REST web services, LLMs, and more.What You'll Do:This role spans the full ML lifecycle for recommendations — from candidate generation through ranking, serving, and online evaluation. Concretely:ModelingDesign, train, and ship recommendation models including two-tower / dual-encoder retrieval, neural ranking, and re-ranking modelsBuild embedding pipelines for users, deals, merchants, and content; iterate on representation learning approachesImprove candidate generation strategies, including ANN-based retrieval over learned embeddingsDefine and run rigorous offline evaluation (recall@k, NDCG, MAP, calibration) and partner with data science to design online A/B testsPartner with product and data science on personalization surfaces — homepage, feeds, deal pages, search re-ranking, and lifecycle channelsInfrastructureBuild and own end-to-end ML pipelines for recommendations: data preparation, training, evaluation, deployment, and monitoringDesign and operate low-latency model serving for high-QPS recommendation trafficBuild feature pipelines and feature-store patterns that maintain online/offline parityDesign, architect, and build reliability, observability, and utilization infrastructure for the recommendations stackImprove training cost, turnaround time, and reproducibility on the ML platform; collaborate with data scientists to unblock experimentationCross-cuttingEncourage change, especially in support of ML engineering best practices, and maintain a high standard of excellenceCollaborate with engineers within the team and across the company to solve complex data problems at scaleWrite high-quality, product-level code that is easy to maintain and test following standard methodologiesWhat We're Looking For:8+ years of relevant professional experienceDemonstrated experience designing, training, and shipping recommendation systems in production — not just classifiers or general MLHands-on experience with deep learning for recsys: two-tower / dual-encoder models, embedding-based retrieval, neural ranking, or similarStrong ML fundamentals: model evaluation methodology, A/B testing, debugging models at scale, handling data and label quality issuesProficiency with ML modeling frameworks (PyTorch and/or TensorFlow) (5+ yrs)Experience with model serving platforms (TorchServe, TensorFlow Serving, NVIDIA Triton, or comparable custom serving infrastructure)Experience with vector retrieval / ANN at scale (e.g., FAISS, ScaNN, OpenSearch k-NN, Pinecone, Weaviate, or similar)Experience working with cloud data processing technologies such as Apache Spark, Elasticsearch, Presto, SQL (3+ yrs)Proficiency in at least two of: Linux, Ansible, Docker, Kubernetes (5+ yrs)Experience in distributed computing (7+ yrs)Experience working with AWS or similar cloud infrastructure (5+ yrs)Experience with hardware / resource management for ML training and/or deploymentKnowledge of the open source landscape with judgment on when to choose open source versus build in-houseExcellent analytical and problem-solving skillsComfort operating across both modeling and infrastructure — this is not a pure modeling or pure platform roleNice to have:Experience with feature stores (Feast, Tecton, or custom)Experience with real-time / streaming feature engineeringExperience with LLM-augmented retrieval or hybrid retrieval architecturesE-commerce, content, or marketplace recommendation domain experienceLOCATION: San Mateo, CAHybrid schedule visiting our San Mateo office three days a week (Tues-Thurs).Slickdeals Compensation, Benefits, Perks:The expected base pay for this role is between $170,000 - $220,000. Pay is based on a number of factors including market location and may vary depending on job-related knowledge, skills, and experience. Exact compensation will be discussed during the interview process and tailored to the candidate's qualifications.Competitive base salary, annual bonus, and equity packageCompetitive paid time off in addition to holiday time offA variety of healthcare insurance plans to give you the best care for your needs401K matching above the industry standardProfessional Development Reimbursement ProgramWork AuthorizationCandidates must be eligible to work in the United States.Slickdeals is an Equal Opportunity Employer; employment is governed on the basis of merit, competence and qualifications and will not be influenced in any manner by race, color, religion, gender (including pregnancy, childbirth, or related medical conditions), national origin/ethnicity, veteran status, disability status, age, sexual orientation, gender identity, marital status, mental or physical disability or any other protected status. Slickdeals will consider qualified applicants with criminal histories consistent with the "Ban the Box" legislation. We may access publicly available information as part of your application.Slickdeals participates in E-Verify. For more information, please refer to E-Verify Participation and Right to Work.Slickdeals does not accept unsolicited resumes from agencies and is not responsible for related fees.