Senior Software Engineer (Machine Learning)
MaropostMatch Score
Against your primary resumeLogin to See Your Match Score
Create a free account or log in to unlock your CV match score across:
Protocol Intelligence
Data-driven signals on your job's competitivenessRemote role, mid-level (5+ years) and popular ML title increase candidate competition significantly.
Medium: core ML and MLOps skills transfer across industries, but recommender and e-commerce specifics increase domain bias.
Explicit 5+ years, 3+ years production serving, and mandatory ML, ANN, vector DB, and Kubernetes skills create high strictness.
Job Description
Structured overview of role & requirementsAbout This Role
Own end-to-end machine learning systems powering product discovery and search ranking for merchants, including personalized recommendations and an LLM shopping assistant.
Design, train, and deploy recommendation and ranking models and build low latency Python services serving live traffic under strict performance budgets (~200ms).
Operate full production lifecycle including batch pipelines, AB testing, containerized deployments on Kubernetes, and performance monitoring and scaling.
Minimum Requirements
5+ years building machine learning systems with at least 3 years serving models in live production.
Strong Python skills including FastAPI or equivalent framework.
Experience with vector databases (like Qdrant) and approximate nearest neighbor libraries such as FAISS, Annoy, ScaNN, or HNSWlib.
Comfortable with containerization (Docker) and Kubernetes deployment; experience running AB tests and strong SQL skills (ClickHouse is a plus).
Ideal Candidate Profile
Hands-on experience with recommender systems or search ranking using implicit feedback, embeddings, and ANN search technologies.
Expertise in operating production machine learning systems independently including building, deployment, testing, and scaling in a small agile team environment.
Experience with LLM applications, e-commerce or marketplace search relevance, and streaming systems like Pulsar or Kafka is advantageous.