Machine Learning Engineer, Ranking & Retrieval

Role overview Develop machine learning systems that help a consumer communication product find relevant information and rank it effectively for each user. The role covers the broader model lifecycle, from retrieval and ranking improvements through evaluation, deployment, and production monitoring.

Responsibilities - Build and refine models for retrieval, ranking, classification, and personalisation. - Improve search relevance and context retrieval with embeddings, hybrid search, and reranking techniques. - Develop ranking and personalisation approaches using user behaviour and product data. - Create evaluation datasets, pipelines, and metrics to assess quality and detect regressions. - Train, fine-tune, deploy, and monitor models in production; investigate issues and iterate using feedback and evaluation results. - Optimise model quality alongside latency, inference cost, and reliability.

Requirements - At least three years of machine learning engineering experience, including production deployments. - Strong Python programming skills and experience with PyTorch or JAX. - Understanding of machine learning algorithms, model training, and evaluation. - Experience with ranking, retrieval, classification, recommendation, or personalisation models. - Practical experience with embeddings, vector search, or information retrieval. - Experience building data pipelines, training workflows, and inference services, and evaluating models with appropriate metrics.

Nice to have - Work with transformer models, language model fine-tuning, or embedding models. - Experience in hybrid search, reranking, and retrieval evaluation. - Recommendation or personalisation experience in consumer products. - Familiarity with large-scale ML training or inference infrastructure.