We are looking for a data scientist who wants to create models and take care of deployment, who is not afraid to speak to client users to get more insights and feedback, and who’s excited to supercharge our Railway predictive maintenance solution with AI/ML alerts, recommendations and insights.
We collect data in real time on running trains and on train maintenance, and turn that data into meaningful information to prevent failures, to reduce delays and downtime, to reduce energy consumption, and to extend maintenance periodicities on railway fleets.
Our datasets include train geolocation, fuel consumption, up to 9000 sensor time series, machine or human failure text information, as well as geographic, traffic, and maintenance data.
If you’re able to access and transform data using usual data extraction techniques (SQL, APIs, parquet, python, batch preparation jobs ...), and if you have some degree of mechanics love in order to find valuable opportunities in a railway dataset, then we’re looking for you! You also apply down-to-earth exploratory techniques and relate to physical models and failure modes and user needs. You design and validate your models and deploy them in production, providing human readable information to end users and collecting feedback from end users (true/false positives, white box explanation...).
Click the button below to apply for this job, and fill out the application form. Please also submit, in the application form, a short notebook exploring a time series dataset of your own choice to show how you explore data, identify value opportunities and how you turn this data into meaningful information / actionable actions for your clients.