About CRU:
CRU International is a leading provider of business intelligence and consulting services in the metals, mining, and fertilizer industries. With over 50 years of experience, we offer valuable insights and analysis that help our clients make informed decisions in an ever-changing global market.
About the role:
CRU Group is seeking a technically strong, hands-on Senior Data Scientist to join a global team focused on building real-world data science and AI solutions for Risk & Trading workflows. This role will span predictive modelling, machine learning, and modern AI systems powered by large language models.
This is an opportunity to own projects end-to-end, from problem framing through to production delivery, working closely with stakeholders, engineers, and technology teams to deliver scalable, product-grade solutions.
Key Responsibilities:
- Build and deliver solutions across predictive analytics, machine learning, and LLM-based AI.
- Design, develop, and implement bespoke ML and AI agents and multi-step workflows using modern frameworks such as LangChain, LangGraph, or similar technologies.
- Develop robust, scalable, production-grade Python code.
- Partner with senior stakeholders to translate business needs into clear technical requirements and measurable outcomes.
- Own delivery end-to-end, including problem framing, data discovery, model development, evaluation, integration, deployment, monitoring, and iteration.
- Collaborate with global teams of analysts, product managers, product modellers, and engineers to deliver high-quality, reliable solutions.
- Mentor team members and contribute to raising standards for technical execution and delivery discipline.
- Stay current with developments in machine learning, statistical learning, generative AI, LLMs, and agentic workflows, and identify practical opportunities to apply these technologies within risk and trading services.
Qualifications:
- MSc or higher in a quantitative discipline such as econometrics, statistics, mathematics, physics, or similar.
- Expert-level Python skills and strong command of the core scientific stack, including Pandas, Scikit-learn, SciPy, and NumPy.
Skills & Experience:
- 3–5+ years of experience in data science, machine learning, econometrics, or a related technical field.
- Proven track record of independently delivering solutions, including model design, training, evaluation, deployment, and monitoring.
- Strong exposure to time-series forecasting or commodity data.
- Creative thinker with a robust approach to data and modelling.
- Hands-on experience working with machine learning and LLMs.
- Hands-on experience building AI applications beyond simple chatbot interfaces.
- Goal-oriented, with a high level of energy and enthusiasm; able to operate efficiently, effectively, and professionally in a fast-paced, dynamic, results-driven business environment.
- Analytical mindset with the ability to quickly understand the intricacies and complexities of commodity data.
- Creative application of first principles to problems, with an awareness of business time constraints.
- Approachable, positive, and collaborative, with a genuine interest in developing the team to be the best in the industry.
- Lead by example in work ethic and commitment to exceeding client expectations.
- Strong commitment to sharing knowledge and ideas in a constructive and supportive way, with the ability to mentor more junior talent.
- Delivery-focused, with a passion for quality and innovation, and a flexible, adaptable attitude willing to accept and drive change.
- Familiarity with Snowflake, AWS, and GCP, or equivalent cloud infrastructure.
- Strong understanding of modern machine learning workflows and production deployment considerations, including orchestration tools.
What We Offer:
- Competitive salary and flexible benefits package.
- Opportunities for professional growth and development as part of a global company.
- A collaborative and supportive work environment.
- The chance to work with industry-leading experts and over a diverse range of topics and projects.
