Embedded AI Engineer

We are looking for an Embedded AI Engineer to build, optimize, and deploy neural networks on ultra-low-power, memory-constrained SoCs and AI accelerators.
You will work at the intersection of AI, embedded systems, and silicon, translating trained models into highly efficient on-device inference pipelines for battery-powered and always-on edge applications for vision, audio, activity and time-series applications.

Tasks & Responsibilities

• Deploy AI models on resource-constrained in-house design SoCs

• Optimize models for memory, power, and latency

• Perform model quantization (INT8 or lower)

• Map neural networks onto microcontrollers, DSPs, and AI accelerators

• Implement inference pipelines in bare metal or Zephyr OS (bonus)

• Validate embedded inference against Python reference models

• Debug numerical mismatches between float and quantized inference

• Optimize memory layout, buffer reuse, and tensor ordering

• Collaborate closely with hardware and firmware teams

Requirements

• Strong fundamentals in machine learning

• Hands-on experience deploying AI on embedded or edge devices

• Proficiency in C/C++ (embedded) and Python

• Experience with quantization (PTQ or QAT)

• Understanding of memory-constrained systems

• Ability to optimize latency, memory footprint, and power consumption

• Experience with TensorFlow Lite Micro, TVM, CMSIS-NN, or ONNX Runtime

• Knowledge of RISC-V architecture or custom AI accelerators

• Experience with edge sensors (audio, IMU, vision, wearables)

• Familiarity with power profiling and optimization

• Agile development experience