Head of Telematics Data Science

About Zego

Insurance hasn’t changed much in over a century. The way we live, work and travel has.

Good drivers are the entire point of Zego. Seven in 10 drivers on the road are consistently good, but until telematics came along there was no way to prove it, so nearly everyone paid a price based on averages. Zego is the company built to change that.

At Zego, we’re leading the AI evolution in insurance.

We move fast, without the layers that usually slow things down. AI takes the routine work off every team, from Finance to Product to Marketing, so people ship real code and launch custom dashboards in minutes. Engineering commits are entirely AI-assisted, which shifts the focus from team size to the quality of your thinking, and our data warehouse connects directly to an AI assistant for instant answers. Backed by a modern tech stack (Claude Code, Snowflake, Figma and Zapier) and a commitment to adopting new tools in days rather than quarters, you get complete ownership to do the most impactful work of your career.

Purpose of the role

Zego has recorded over a billion miles of real-world driving, captured from phone sensors (accelerometer, GPS and more) across tens of thousands of policies and four very different kinds of driving: private car, van, taxi and delivery. The Telematics team uses digital signal processing and machine learning to turn that raw signal into a precise picture of how people actually drive. We use it to price insurance more fairly, to help drivers be safer on the road, and to power Zego Rewards, everyday rewards for good driving that turn car insurance into an engaging product rather than a box to tick.

We own the full telematics stack (mobile SDK, on-device ML, cloud pipeline) and the full pricing stack. So we can link driving behaviour directly to claim type, frequency and severity, and feed one system's predictions straight into the other. We are one of very few companies that owns that entire loop end to end, on in-house proprietary technology.

As Head of Telematics Data Science you will set the research agenda and own the technical direction of that work, reporting to the Director of Data. You will lead three data scientists directly and three ML engineers with a dotted line to you. The split is roughly 60% leading, 40% hands-on.

Our long-term ambition is driving behaviour modelling that holds up across countries and handsets, is highly predictive of claims risk, and is deeply integrated with driver coaching so we can offer the fairest price and the best experience to good drivers anywhere. Our partnership with Sompo Holdings, one of Japan's largest insurance groups, will take models trained on UK roads into an entirely different market.

The problems you will be working on

Continuously pushing the predictive power of our existing algorithms and ML models

Telling drivers apart from passengers using sensor data alone

Detecting crashes and estimating their severity

Bringing map and road-context data into our models in a way that genuinely improves prediction

Handling handset and sensor heterogeneity, so a model behaves consistently on a five-year-old Android and a new iPhone

On-device ML under real battery and compute constraints - trip detection already runs on the phone, and more could

Making behavioural risk models transfer to new markets, starting with Japan

Several of these are open problems across the industry, not just at Zego.

What you will be doing

Define and drive the technical roadmap, making sure the team is working on the highest-impact problems

Lead the development of novel models to detect human behaviour and vehicle kinematics from high-frequency phone sensor data, on-device and in the cloud

Act as the primary technical authority on telematics across the business, advising on existing products, surfacing patterns nobody has thought to look for, and identifying new value propositions

Build, grow and mentor the team, setting the technical standards needed to produce e