Samsung Knox is looking for a Research Engineer Intern to research innovative ideas in the cyber security domain and create ML solutions for Samsung's array of B2B products, aiming to integrate these solutions into their Security & Intelligence services in Production.
Requirements
- Familiar with AI/ML latest framework, such as PyTorch or TensorFlow
- Strong hands-on experience in threat detection
- Strong hands-on experience in leveraging heterogeneous tabular data for feature learning, graph construction, deep graph learning, weak supervision, causal reasoning, structure learning, out-of-distribution quantification, reinforcement learning and/or knowledge distillation is appreciated
- Strong hands-on experience in conducting architectural changes on given ML solutions inspired by research papers and/or technical discussions
- High-quality coding in Python
- Familiar with optimization processes for ExecuTorch, ONNX and/or LiteRT
- Familiar with ML model deployment and inferencing in Android OS
- Familiar with Data Structures, Algorithms and Linux OS concepts
Responsibilities
- Create quick prototypes and proof-of-concepts
- Explore cutting-edge technologies for our products
- Design, develop, and productize AI/ML based solutions for Samsung services
- Design, develop, and productize Machine Learning algorithms from Samsung Research teams or research papers
- Design experiments, perform evaluations, and apply enhancements to our products
- Review work with the engineering team weekly and able to present your solution to SRA by the end of your internship
Other
- Ph. D. or M.S. students in Computer Science, Electrical & Computer Engineering or relevant fields
- Open communicator with excellent visual, verbal, and written communication skills
- Passionate about researching innovative solutions and creating enterprise products for web and mobile
- First- or co-authored publications at leading workshops or conferences such as NeurIPS, ICML, ICLR, CVPR or similar.
- This position will be performed in an office setting.
