- Data Science
Predictive modelling for car damage image recognition, at CarNext.com Amsterdam
For each of LeasePlan’s ex-lease cars (250,000 per year), an inspector performs a visual inspection and makes a report to assess the damage on the car. Photos and report are currently manually evaluated by a CarNext.com expert to determine if the quality of the car is high enough for selling it B2C in one of CarNext.com’s stores, i.e. if damages on a car are small enough to ‘smart repair’.
Goal of the project is to build a predictive model and an image recognition engine to select the best cars for CarNext.com’s stores in NL. You will be writing algorithms and apply machine learning for this purpose. You will be supervised by an experienced Data Scientist and you will have frequent contact with the damage experts to acquire practical information. Next to this, you will be part of a young, fast-growing, international and enthusiastic team.
If successful, you will have a huge impact: the model will be implemented in 20+ countries in Europe.
We are looking for someone who
- Has advanced Python skills
- Is pragmatic and has proven that she/he can deliver
- Has some experience with machine learning techniques.
- Is currently enrolled in a science-, engineering- or econometrics-related Master/Bachelor program.
- Is available for at least 3 days/week and 3 months
Students can apply by sending an email with their resumé and availability to Karim Bennani, Data Scientist at email@example.com.