The job ad is older than 1 month and may no longer
We are looking for a full-time senior scientist or an engineer to perform AI Speech/Audio research. At Samsung SDS Research Center, we use AI and deep learning technologies to assist global companies with their digital transformation.
This position will require you to conduct speech research such as speech enhancement using deep learning techniques. Your knowledge will be used to support theoretical and experimental research, as well as to improve video conferencing and collaboration solutions in the B2B field.
Samsung SDS America (SDSA) is the U.S. subsidiary of Samsung SDS, a global software solutions and IT services company. Samsung SDSA helps organizations optimize their productivity and make smarter business decisions using our enterprise software solutions for secure mobility, HPC managed services, DOOH, and advanced analytics and contextual marketing. We enable our customers in government, financial services, retail, and other industries to drive business and improve their competitive positions in a hyper-connected economy.
This position may be based in San Jose, CA (preferred) or remotely in the U.S. (except Colorado).
- Conduct speech enhancement research such as noise suppression, echo cancellation, and double-talk detection with a global team.
- Use deep learning technology for speech research – you may be engaged in developing a hybrid model of signal processing and deep learning.
- Develop AI models and conduct theoretical and experimental research. You will also be responsible for documenting your work, writing a paper, and filing a patent application.
- Use frameworks such as ONNX, TensorRT, OpenVINO to optimize and compress AI models for field application.
- Use programing languages such as Python, Java, and C++ to make the development, mainly on a dockerized Linux platform in the cloud. (However, a closed network environment may need to be used due to customer security issue.)
- Develop AI models and perform experiments using pyTorch, Tensorflow, and more.
- Collaborate with multinational researchers.
- Degree in relevant fields including (but not limited to) computer science, electrical engineering, mathematics, statistics, and physics. (PhD preferred)
- Published papers, whitepapers, and patents related to speech processing (e.g., deep learning-based speech enhancement).
- 5+ years of experience with strong, demonstrated programming skills in at least one of the following languages: Python, Java, C, C++.
- 5+ years of experience and knowledge in acoustic signal processing.
- Good communication and presentation skills in working with multinational researchers.
- Creative problem-solving skills and planning skills for research topics.
- Ability to travel to HQ in Korea ~2x/year
- If Remote, ability to travel to San Jose office as needed
- Experience in software-based acoustic signal processing.
- Research experience in deep learning related to self-supervised learning methodology, RNN, and transformer system.
- Field experience in supporting solutionization or commercialization of research models.
- Experience in areas such as deep learning model compression and high-speed inference or ability to port deep learning models to C++.
- Experience building Docker/Kubernetes environment and developing MSA services.
- Experience with top-tier conference presentations or publication of papers in high impact factor journals.
- Experience in presenting creative research topics and writing research proposals.
- Experience working with multinational research teams.
Samsung SDS America, Inc. is an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to age, race, color, religion, sex, sexual orientation, gender identity or expression, national origin, disability, status as a protected veteran, marital status, genetic information, medical condition, or any other characteristic protected by law.
Benefits: Medical, Dental, Vision, Prescription, 401k, Paid Time Off (PTO)
Must be capable of working in San Jose, CA or remotely from the U.S. This role cannot be performed in Colorado