Machine Learning Engineer

About the Role

This position places you in the Platform team, responsible for maintaining and evolving the backend platform integral to Xantura's core operations. You will manage a predictive modelling platform designed to address a variety of challenges across different problem domains and clients, including predicting vulnerabilities related to housing, health, and social issues.

Key Responsibilities

• Develop and improve scalable predictive models such as embedding-based sequence encoders, temporal survival models, and gradient boosted decision trees.
• Stay current with the latest advancements in machine learning and frontier models, conducting rigorous experiments to safely integrate new techniques into production.
• Create and maintain robust evaluation frameworks, training datasets, and model infrastructure to continually enhance natural language processing and predictive analytics capabilities.
• Ensure all AI deployments adhere to ethical standards and regulatory requirements throughout the development process.

Candidate Profile and Requirements

• Possess a Bachelor’s or Master’s degree in Computer Science, Machine Learning, or a related discipline, or have equivalent hands-on experience.
• Minimum of three years' professional experience in machine learning engineering or a similar role.
• Strong expertise in Python programming with practical experience deploying code in production environments.
• Experience with orchestrating data or machine learning pipelines using tools like Dagster, Airflow, or Prefect.
• Proficient in popular ML frameworks and libraries, including PyTorch, scikit-learn, and gradient boosting implementations such as XGBoost or LightGBM.
• Demonstrated ability to define and deploy containerized applications, including API implementation (e.g., via FastAPI) and production deployments using Kubernetes.

Preferred Additional Skills

• PhD qualification with a strong publication history in areas such as text analytics, representation learning, or applied predictive modeling.
• Experience working with vector databases and building retrieval-augmented generation (RAG) pipelines, including setup and configuration of vector databases.
• Familiarity with deploying recent AI models through platforms like Huggingface transformers and OpenAI APIs.
• Development of agentic systems using tools such as LangChain, AutoGen, or PydanticAI.
• Active involvement in open source projects, hackathons, or possessing shareable programming work.
• In-depth knowledge of embedding architectures including bi-encoders and cross encoders, especially for long-term text or temporal tasks.
• Experience in constructing and delivering asynchronous, production-grade APIs for embedding or heavy computational services.
• Strong software engineering practices, including testing, version control, and continuous integration/delivery for Python-based data and model pipelines.
• Good command of Microsoft Azure services, including Azure Kubernetes Service, Azure Batch, Azure AI Foundry, and Azure Machine Learning components.

Location and Work Arrangement

This role is primarily hybrid, based in the London office, requiring presence onsite a minimum of 1–2 days per week.

About Xantura

Xantura focuses on addressing societal inequality by enabling local authorities to utilize data more efficiently. The company’s AI-powered platform integrates various siloed datasets, applies advanced machine learning techniques, and offers predictive analytics to identify at-risk individuals before critical social issues arise. This generates actionable insights aiding frontline workers in early intervention to improve outcomes related to homelessness, child welfare, and other social challenges. Xantura is experiencing rapid growth, expanding its client base and technological platform with a mission centered on improving lives through data and AI.