Machine Learning Engineer

Who we are
OTIV is a Ghent-based company leading the autonomous revolution in rail. We are on a mission to increase safety and efficiency of railway operations by teaching rail vehicles to drive autonomously. By making trains more efficient and safer, we actively contribute to a modal shift and a greener future. We bring this vision to life by a unique focus on the most challenging environments such as cities, shunting yards and industrial sites. Thanks to high-profile partners and world-leading clients such as Deutsche Bahn, SNCF Group, ArcelorMittal and CAF we deliver the future of railway operations today.
OTIV is an equal opportunity employer. We are committed to a work environment that supports, inspires, and respects all individuals and in which personnel processes are merit-based and applied without discrimination on the basis of race, color, religion, sex, sexual orientation, gender identity, marital status, age, disability, national or ethnic origin, military service status, citizenship, or other protected characteristic. OTIV strives to create an environment in which all employees feel confident in their ability to bring their authentic selves to work and secure collaborating with colleagues and customers.

What you will do
As an experienced ML (Machine Learning) engineer, you will research, train, and validate neural networks operating on camera images and/or 3D point clouds. You will contribute to the main AI component of ADAS (Advanced Driver Assistance Systems) and self-driving systems. Since these models are trained on OTIV's own dataset, you will be responsible for maintaining the datasets used for training and validating models. You will work on a combination of end-to-end multi input neural networks and single data source neural networks.

Tasks and responsibilities

• Research & selection of neural networks for a variety of tasks: object detection, rail detection, etc..
• Training & validation of the selected models
• Create, organize and take ownership of datasets used for training and validation
• Deploy AI models on edge devices
• Supervise third party annotation teams
• Closely collaborate with data engineers and robotics engineers

Requirements
Who you are

• You are experienced in end-to-end neural networks for vision based applications: you know how to train, finetune, scale and deploy them.
• Master/PhD in computer engineering, AI or equivalent
• Strong coding skills in Python
• Proven ability to train and validate neural networks in Pytorch (using ONNX)
• Experience in handling and maintaining large datasets
• Experience in solving computer vision problems
• Knowledge of CUDA and/or TensorRT
• Fluent in English and knowledge of any other European language is a plus.

Benefits
What we offer

• Full time offer with competitive salary package.
• International environment in a disruptive scale-up with opportunities to grow.
• Early riser or night owl? Either way, you’ll fit right in our flexible working hours system.
• Office-first culture, with an office in beautiful Ghent (Belgium).
• Collaborative, transparent, and fun-loving office culture.
• Team events and quarterly offsite.
• 32 days of vacation.