Full-time

Data Scientist, Telematics

London or Riyadh, Remote considered (Hybrid)  ·  Data Science  ·  Posted 5 November 2026  ·  Open until filled

Apply now

Role summary

Reports to
CTO
Territory
Global
Location
London or Riyadh, Remote considered · Hybrid
Focus
Every second, our vehicles and sensors produce a torrent of data. Your job is to turn it into decisions that keep drivers safe, cut insurance losses and save fleets real money.
Measured on
Successfully own the full journey from raw signal to shipped product. Your models are trusted. They will accurately modify behaviour, reducing accidents and save lives. Your models go beyond reacting to predicting.
Experience
4+ yrs of data science including hands-on exposure to telematics data, and you’re comfortable working with noisy, high-volume time-series data.
Languages
English
Start date
As soon as you can join us

Real, live data at scale, not a static dataset. Your models ship, and you’ll see their impact in customers’ hands. Real ownership of the data science roadmap

About the role

  • You'll develop models that are accurate, trusted and predictive. 
  • You are not just doing a job, you are growing a business.

About you

  • In addition to direct experience, you'll be excited to join a lean start-up working with small fast teams where your ideas get built rather than queued. 
  • You will be excited to have broad company involvement, autonomy and impact.  You'll be a Big fish in a small pond developing amazing models.
  • You'll be a self-starter, clever and innovative
  • You'll be able to translate complex ideas communicating them to others.

In this role, you will

  • Every second, our vehicles and sensors produce a torrent of data. Your job is to turn it into decisions that keep drivers (and other road users) safe, cut insurance losses and save fleets real money. 

You’ll own the full journey from raw signal to shipped product:

Teach the sensors to see what really happened.

You’ll build driving behaviour event detection from the ground up, including the calibration work that makes a harsh brake or a swerve mean the same thing in a van, a truck or a hatchback, however the device was fitted. Get this right and everything downstream is trustworthy.

Create the risk score the industry relies on

. You’ll model accident frequency and severity from real driving behaviour and distil it into a single score that insurers will price on and fleet managers will act on. This is the model that changes behaviour and premiums.

Predict what’s next.

Fuel burn, maintenance failures, customer lifetime value and other high-impact use cases are all on the table. You’ll take each one from first prototype to production, with no hand-off to another team and no model left gathering dust in a notebook.

Minimum requirements

  • You will hold an MSc or PhD in a quantitative subject such as statistics, mathematics, physics, computer science or engineering.
  •  You are able to combine rigorous statistical and machine learning thinking with a pragmatic focus on models that perform reliably in the real world, and you can explain your work clearly to technical and non-technical audiences alike.
  • 4+ years relevant experience
  • Fluent English, written and spoken. 

An ideal candidate also has

  • Ideally you will have knowledge and experience of motor insurance and claims analytics.

Rooya is an equal-opportunity employer. We welcome applications from everyone regardless of background.

Don’t see the right role?

We’re always glad to hear from people who break the mould. Send us your CV, and we’ll keep it on file.

    No commitment — we’ll only be in touch about roles that fit.