Senior Embedded software engineer

Our client is situated in the Eindhoven area. They are responsible for new innovations in eg the automotive, industrial & IoT and mobile markets.

Your main goals will be to investigate, design, integrate, and test state of the art NAS in cooperation with team-members. The objective is to:

• develop a suitable NAS methodology

• prove and demonstrate the benefits to make it a strategic pillar in AI solution portfolio

• provide external thought leadership and marketing communication with the developed solutions and application examples.

Responsibilities

• Investigation state of the art approaches in (hardware-aware) Neural Architecture Search.

• Exploration, design, prototyping NAS-methods in conjunction with selected automated driving related NN-based applications and evaluation of the methodologies.

• Profiling of NN operations for various company’s processors.

• Interaction with external application partners.

Experienced required

• A university degree: MSc, PhD, with significant exposure/involvement with Machine Learning or neural networks. Experience with NAS methodologies is a pre but not required.

• Stellar results during the MSc. or PhD., or more than 5 years of professional experience in industry or academia.

• Affinity and experience with embedded processors, software and NN accelerators.

• Broad experience with embedded software architectures, build systems, version control systems.

• Broad experience with Operating systems GNU/Linux, embedded systems, development boards, and processors, and SW competencies as listed below

• AI research, development and debugging experience

• Translating AI research into AI deployment

• Flexibility in working with AI frameworks (TensorFlow, PyTorch), preferably via Python and C++ interfaces

• Familiarity with setting up and maintaining related development environments (Jupyter notebooks, TensorBoard, profiling, docker, etc.). Understanding of AI toolchains, deployment, portability and inference engines (CUDA, TensorRT, TFLite, ONNX, etc.)

• Ability to value, interpret and prioritize AI s.o.t.a. in terms of realistic real-world conditions and constraints