ML-Ops Engineer

About The Position

Samsung R&D Center is looking for an ML-Ops Engineer to join our team.

Samsung (SIRC), shaping the world of tomorrow, Today. Focusing beyond the horizon and pushing exciting developments in many key areas of technology. Samsung is creating a new era of continuous innovation, bringing value and contribution to society and creating a workplace where our employees can enjoy making the most of their talent, creativity and passion.

The team:

SIRC’s Sensor Team is a world class R&D group that specializes in building the next generation of cameras for the mobile, automotive and consumer markets.

The sensor algorithms group is leading the design and implementation of cutting-edge image processing and computer vision solutions aiming at improved quality and extended capabilities of image sensors.

What will you do?

• Work closely with algorithms teams to build machine learning infrastructure.
• Take responsibility for end-to-end design of APIs, methodologies, processing pipelines and integration with internal customers and tools.
• Help establish best practices and a culture of excellence in deploying maintainable data and model pipelines.
  
  

• B.Sc. in computer science or related STEM degree.
• At least 5 years SW industry experience including MLOps roles.
• High proficiency in Python and C++.
• Strong understanding of the MLOps technological landscape, software testing methodologies and continuous integration tools.
• Strong communication and interpersonal skills.
• Ability to effectively communicate across multi-disciplinary teams, translate discussions into robust software tools, designs and guidelines.
  
  

• Hands-on experience with low-level internals of machine learning frameworks including quantization and deployment of machine learning models to embedded platforms and setup of distributed systems for training of ML models.
• Experience with Pytorch/cmake/docker/git/Linux/GPUs/quantization/job-schedulers and CI/workflow tools.
• Experience with web applications.
• Able to profile performance issues.