Embedded Systems

ML Engineer

Work Type: Full Time

Seeking a Senior software Engineer with extensive experience in developing edge AI and TinyML solutions. The position involves a significant level of autonomy and responsibility in which a self-motivated applicant must work independently to develop solutions.

QUALIFICATIONS / EXPERIENCE

Applicant should possess a Bachelors or Masters in EE or CS.

Applicant must have at least 5+ years of proven experience in developing embedded applications, Tiny ML and edge AI solutions with good platform optimization skills.

Applicant must have experience in several of the following:

Required:

Breadth and depth of knowledge on Computer Vision

Experience on machine learning algorithms, DL frameworks and compilers, such as PyTorch, TensorFlow, TensorFlow Lite, TVM, TFLM, Glow, ONNX

Experience building machine learning solutions applied to real-world problems

Experience working with SoCs, profiling applications and employing various machine learning optimization techniques

Experience of writing Custom Operators for AI Accelerators

Publications in peer-reviewed journals and conferences in relevant fields, such as ICML, ICLR, NeurIPS and CPVR

  • Development of applications in a Linux environment
  • Knowledge of programming using Hexagon DSP vector extensions (HVX), OpenCL

Preferred:

  • Proficient in C99, C++14,  Python
  • Experience with PyTorch, OpenCV, TVM, Glow
  • Experience in writing secure code
  • Experience in all stage of the Secure SDLC
  • Working knowledge of GIT

Job Responsibilities

Applicant will be required to:

Collaborate with a cross-disciplinary team to design and implement novel AI solutions for edge devices

Research state-of-the-art modelling and training techniques

Propose, prototype, and validate hypotheses to lead our business and product roadmaps

  • Evaluate new technologies and innovate to improve product competitiveness
  • Mentor junior engineers and guide them as needed
Liquid error: undefined method `public_fields' for nil:NilClass