Senior MLOps Engineer

About the role

We are looking for a Senior MLOps Engineer to join Elsevier’s Data Science Life Sciences team in Amsterdam. You will bridge data science and engineering, turning experimental NLP, IR and GenAI models into secure, reliable, and scalable services that power our life‑science products such as Reaxys, Embase and Pharmapendium.

Key responsibilities

• Design, build and automate end‑to‑end ML pipelines on major cloud platforms (AWS, Azure, Databricks) and foundation‑model APIs (OpenAI, Bedrock).
• Maintain model registries, artifact stores and versioning to ensure reproducibility and governance.
• Develop CI/CD workflows for ML, including automated data validation, model testing and deployment using tools such as SageMaker, MLflow and Azure ML.
• Implement and optimise recommendation, search and retrieval systems (Elasticsearch, OpenSearch, Solr, vector and graph databases).
• Build evaluation pipelines with offline IR metrics (NDCG, MAP, MRR) and LLM quality metrics (faithfulness, grounding) and run A/B tests.
• Monitor infrastructure costs and apply scaling strategies to maximise efficiency.
• Collaborate with data scientists, product managers, responsible‑AI experts and operations engineers to translate business problems into production‑ready solutions.

Required profile

• 5+ years of experience in ML engineering, MLOps platforms and shipping ML or search/GenAI systems to production.
• Strong programming skills in Python and at least one of Java or Scala.
• Hands‑on experience with major cloud providers (AWS, Azure, Google Cloud).
• Familiarity with statistical analysis, machine learning theory and natural language processing.
• Background in scholarly publishing workflows, bibliometrics or citation graphs is a plus.

Required skills

• Python
• Java or Scala
• AWS (SageMaker, Bedrock)
• Azure (Azure ML)
• Databricks
• MLflow
• CI/CD for ML
• Elasticsearch / OpenSearch / Solr
• Vector databases
• Neo4j (graph DB)
• OpenAI APIs
• Model registry and artifact management
• Evaluation metrics (NDCG, MAP, MRR, LLM faithfulness)