Роль четко определена с акцентом на передовые технологии ИИ, а зарплата соответствует рыночным ожиданиям, что делает эту вакансию привлекательной.
Высокая зарплата
Четкое определение роли
Передовые технологии
Продуктовая компания
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Embedded AI Engineer (Edge AI): Developing and optimizing custom kernels, operators, and runtime components that bring Deepgram speech models to non-NVIDIA accelerators, embedded SoCs, DSPs, and NPUs with an accent on quantization, architecture-specific compilation, and low-level performance tuning. Focus on meeting latency, memory, power, and thermal constraints, integrating vendor toolchains, and validating production inference across constrained hardware.
Location: USA — Remote
Base salary: $219.3K–$274.1K annually, excluding bonus, equity, and benefits.
Company
Deepgram provides real-time speech-to-text, text-to-speech, and voice AI APIs and deployable foundation models for large-scale production use.
What you will do
Write and optimize custom kernels and operators in C, C++, Rust, and platform-specific assembly or intrinsics.
Optimize speech models for non-NVIDIA accelerators, embedded SoCs, DSPs, NPUs, and purpose-built devices.
Apply quantization, operator fusion, memory-layout optimization, and architecture-specific compilation.
Integrate vendor NPU/DSP toolchains and edge inference runtimes, extending them with custom operators.
Build runtime components for embedded Linux, bare-metal, and RTOS environments.
Benchmark and validate latency, accuracy, power, memory footprint, utilization, and regressions across target platforms.
Requirements
Production experience with resource-constrained embedded, mobile, or edge AI systems.
Strong proficiency in C, C++, and/or Rust for performance-critical environments.
Hands-on experience with on-device model optimization, including quantization, pruning, knowledge distillation, or architecture-specific compilation.
Familiarity with edge inference runtimes such as ONNX Runtime, TensorRT, TFLite, or ExecuTorch, and/or vendor NPU/DSP toolchains.
Understanding of CPU, GPU, NPU, and DSP architectures, memory hierarchies, fixed-point arithmetic, and power management.
Experience with bare-metal, RTOS, embedded Linux, microcontrollers, or edge SoC development.
Custom quantization, mixed-precision inference, neural architecture search, or model compilation toolchains.
Hardware evaluation and benchmarking across accelerators, SoCs, or GPUs.
Experience shipping AI features in consumer products or implementing secure on-device deployment.
Culture & Benefits
AI-first operating model with active use and experimentation with advanced AI tools.
Fast-changing environment focused on experimentation, adaptation, and continuous learning.
Compensation includes annual bonus and equity in addition to base salary.
Будьте осторожны
Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →