Netradyne
About this role
Where Netradyne's research roles build the driver monitoring models, this one makes them run fast enough on the hardware in the vehicle. The job is system optimisation: crafting tools, frameworks and reporting for performance, streamlining software for deployment across IoT devices using GPU, CPU and DSP, and optimising the machine learning models already deployed on the platform. You would also improve the driver monitoring and assistance algorithms themselves to boost efficiency, report the key performance indicators, and handle production inquiries and application stability in live environments. The required toolchain is unusually specific and worth reading before applying: C and C++, OpenGL, CUDA and Python, plus ML frameworks including Caffe, TensorRT, OpenCL, SNPE, OpenVINO and ONNX. Experience with embedded platforms, make files and build systems is called out, with Jenkins described as a valuable asset.
Who this is for
What the posting requires:
- A B.E, B.Tech, M.E or M.Tech degree with a minimum of 6+ years of experience in software system optimisation.
- Proficiency in C and C++, OpenGL, CUDA and Python.
- A solid grasp of basic statistics, probability, and machine learning and computer vision concepts.
- Experience with ML frameworks including Caffe, TensorRT, OpenCL, SNPE, OpenVINO and ONNX.
- Exceptional attention to detail, strong analytical skills and a creative mindset directed at system performance.
Described as a valuable asset rather than required:
- Experience with embedded platforms, make files and build systems, and familiarity with Jenkins.
What the work actually looks like:
- Craft and develop tools, frameworks and reporting mechanisms for system optimisation.
- Streamline software for deployment across IoT devices, implementing on GPU, CPU and DSP.
- Enhance data platforms and optimise the machine learning models deployed on the platform.
- Improve driver monitoring and driver assistance algorithms to boost system efficiency, and report the key performance indicators.
- Manage production inquiries and ensure overall application stability in production environments.
- Convey highly technical results to diverse audiences.
Location and working pattern: Bangalore. Office days are not stated in this posting.
A note on the employer: as with Netradyne's other listings, the posting carries a recruitment fraud warning. Official mail comes only from netradyne.com or us-greenhouse-mail.io addresses, and the company never asks candidates for fees or for sensitive data over unsecured channels.
Honest fit guidance: this is the inference optimisation and embedded deployment end of ML, not model research. If your background is training models in Python and you have never profiled something on a DSP or converted a model through SNPE or TensorRT, the framework list is the honest gap. If you have done that work, it is one of the clearer matches on today's board.
- A B.E, B.Tech, M.E or M.Tech degree with a minimum of 6+ years of experience in software system optimisation.
- Proficiency in C and C++, OpenGL, CUDA and Python.
- A solid grasp of basic statistics, probability, and machine learning and computer vision concepts.
- Experience with ML frameworks including Caffe, TensorRT, OpenCL, SNPE, OpenVINO and ONNX.
- Exceptional attention to detail, strong analytical skills and a creative mindset directed at system performance.
Described as a valuable asset rather than required:
- Experience with embedded platforms, make files and build systems, and familiarity with Jenkins.
What the work actually looks like:
- Craft and develop tools, frameworks and reporting mechanisms for system optimisation.
- Streamline software for deployment across IoT devices, implementing on GPU, CPU and DSP.
- Enhance data platforms and optimise the machine learning models deployed on the platform.
- Improve driver monitoring and driver assistance algorithms to boost system efficiency, and report the key performance indicators.
- Manage production inquiries and ensure overall application stability in production environments.
- Convey highly technical results to diverse audiences.
Location and working pattern: Bangalore. Office days are not stated in this posting.
A note on the employer: as with Netradyne's other listings, the posting carries a recruitment fraud warning. Official mail comes only from netradyne.com or us-greenhouse-mail.io addresses, and the company never asks candidates for fees or for sensitive data over unsecured channels.
Honest fit guidance: this is the inference optimisation and embedded deployment end of ML, not model research. If your background is training models in Python and you have never profiled something on a DSP or converted a model through SNPE or TensorRT, the framework list is the honest gap. If you have done that work, it is one of the clearer matches on today's board.
Apply on company site
Opens www.netradyne.com, the employer's own application page. Applying is always free.