Data Scientist, Fleet Operations

THE ROLE
As a Data Scientist in the Fleet Systems and Insights team, you will play a critical role in optimising fleet operations through data-driven insights and operational research. You’ll help identify high-impact opportunities and guide strategic decision-making, driving improvements across the on-road testing lifecycle.
Rather than focusing solely on black-box models, this role emphasizes using operational research techniques, experimental methods, and causal inference to derive actionable insights for operational efficiency and optimisation.
This means you might:
– Develop frameworks to synthesize complex operational data (e.g., vehicle performance, route optimisation, and experiments scheduling) to inform strategy at both the product and company level.

– Identify key performance metrics for fleet operations and continuously refine them to ensure they align with wider business goals.

– Create and apply novel experimental methodologies to enhance the signal-to-noise ratio and speed up feedback loops, improving operational decision-making and optimising use of on-road testing for  ML advancements.

– Combine experimental methods with causal inference techniques to test and optimise operational strategies.

WHAT WE ARE LOOKING FOR
ESSENTIAL
– 3+ years of experience in a Data Science role, with a focus on operations research, process automation and optimisation, or similar fields.

– Proficient in querying and building large datasets, writing production-level SQL for data transformation pipelines.

– Experience designing and evaluating real-world experiments (e.g., A/B testing) to optimize operations and performance.

– Solid understanding of statistical principles, including hypothesis testing, distributions, and assumptions behind statistical methods.

– Proficient in using a statistical scripting language (e.g., Python, R) and relevant packages (e.g., pandas, sklearn, statsmodels).

– Strong ability to summarise, visualise, and communicate data insights in a clear and compelling manner.

– Proven track record of driving operational improvements and influencing team strategies with data-driven findings.

– A focus on actionable insights that can directly inform fleet operations prioritization and optimization strategies.

DESIRED
– Practical experience with machine learning and optimization techniques (e.g., pytorch, scikit-learn).

– Experience promoting statistical rigor and experimental best practices in previous roles.

– Familiarity with causal inference, econometrics, or Bayesian methods for testing hypotheses in operations research.

– Prior experience working with large datasets and distributed computing (e.g., Spark, Hadoop).

– Experience in a fast-paced tech or startup environment.

This is a full-time role based in our office in London. At Keystone AI we want the best of all worlds so we operate a hybrid working policy that combines time together in our offices and workshops to fuel innovation, culture, relationships and learning, and time spent working from home.