Machine Learning Developer (Budapest, HU)

Tasks

• Development of machine learning and computer vision solutions for vehicle recognition and traffic applications.
• Design, implement, teach and evaluate models and data processing processes.
• Develop and use MLOps pipelines: reproducible training and evaluation pipelines, experiment tracking, automation, and reliable delivery of models from development to use.
• Detection of data quality and model errors, design of measurements and experiments.
• Further develop and optimize existing neural network models and try new approaches.
• Collaborate with C++/edge inference and data platform developers to make the models work in a real product.
• Independently carrying out one's own tasks: understanding the problem, proposing solutions, implementation, measuring, documenting and presenting results.

Requirements

• Solid Python knowledge and basic software development skills.
• Knowledge of the basics of machine learning / deep learning; Understanding the teaching and evaluation of neural networks.
• Independent and structured work: is able to understand the purpose of the task entrusted to him/her, plan the necessary steps and complete the task with appropriate professional consultations.
• Initiative and responsibility: actively seeks solutions, formulates his own proposals, and takes responsibility for the progress and results of the tasks entrusted to him.
• Motivation for development and learning, openness to learn new technologies, models and tools independently.
• Cooperative and transparent communication: regularly and clearly indicates progress, uncertainties and possible obstacles, and asks for help in time if necessary.
• Basic Git usage and openness to development in teams.

An advantage

• Computer vision projects or university/own ML project.
• Practical knowledge of TensorFlow/Keras, PyTorch or similar framework.
• C++ knowledge or interest in production or edge inference.
• Knowledge of Docker, Linux, Kubernetes, or MLOps tools.
• Experience with model optimization, quantization, TFLite or edge tools.
• Experience in data processing, measurement methods, or model evaluation.
• Your own project, TDK, thesis, competition or open-source work, where you carried out a problem independently.

What we offer

You can work on real-world, industrial computer vision and machine learning problems, on large amounts of real data, on models that are used in working products.

The work doesn't end with training a model: from understanding data and designing experiments, to model development, MLOps processes, and optimization, to production and edge, you can follow the lifecycle of a solution.

The depth of the tasks and the degree of autonomy are adapted to your experience. As a career starter, you will receive support for professional development, and with a few years of relevant experience, you will have the opportunity to carry out more complex problems on your own, develop technical solutions and take over greater professional responsibility in a particular area.

You will work in a small, experienced team of engineers where your own ideas and professional suggestions are counted on. We are looking for the implementation of non-finished recipes: it is important for us that you understand the problem, consider the possibilities and actively shape the solution.

We don't expect you to be familiar with all of the technologies listed. It is more important for us to have a stable technical foundation, curiosity, willingness to learn, and to love understanding a problem and taking it to a real result.

Please note: This job description is translated from Hungarian. 

 

Apply for this job