Назад
Company hidden
4 часа назад

Perception Engineer (Computer Vision)

125 000 - 160 000$
Формат работы
onsite
Тип работы
fulltime
Грейд
senior
Английский
b2
Страна
US
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Perception Engineer (Computer Vision) (AI/autonomous defense): Designing, training, and deploying vision and multi-sensor perception systems for navigation, targeting, and automatic target recognition with an accent on real-time embedded inference, EO/IR imagery, and robust model performance in degraded environments. Focus on building detection, segmentation, and tracking models, fusing radar/LiDAR/RF inputs, optimizing CUDA-based deployments on Jetson-class hardware, and validating systems with simulation and flight data.

Location: On-site in Huntington Beach or San Francisco, California, United States

Salary: $125,000–$160,000 base annually for Perception Engineer; $165,000–$220,000 base annually for Senior Perception Engineer, plus equity.

Company

hirify.global is a startup defense technology company founded in 2023 that develops autonomous, decentralized defense platforms and AI-based autonomy systems for contested environments.

What you will do

  • Design, train, and deploy computer vision and multi-sensor perception systems for navigation, targeting, and automatic target recognition.
  • Build detection, segmentation, and tracking architectures for EO/IR and multispectral imagery, including large-scale datasets and training pipelines.
  • Develop evaluation, regression testing, model-update, drift-monitoring, and robustness workflows using simulation and real-world field data.
  • Optimize models for real-time embedded inference on Jetson-class hardware using quantization, pruning, TensorRT, ONNX Runtime, and CUDA profiling.
  • Fuse vision outputs with radar, LiDAR, and RF cues for confirmation, association, and track management.
  • Create visualization, triage, root-cause analysis, and end-to-end test tools for simulation, hardware-in-the-loop, and flight testing.

Requirements

  • 7+ years of relevant experience and a BS, MS, PhD, or equivalent experience in computer science, electrical engineering, robotics, or a similar field.
  • Production C++ on Linux and Python for machine learning and tooling, with rigorous profiling, optimization, testing, and documentation practices.
  • Strong experience with modern detection, segmentation, and tracking models and training or fine-tuning in PyTorch.
  • Experience building diverse datasets, labeling and QA pipelines, augmentation workflows, experiment tracking, and reproducible training systems.
  • Hands-on experience with INT8/FP16 model compression, runtime optimization, real-time constraints, EO/IR imagery, and challenging flight or test data.
  • Must be authorized to work in the United States and may need authorization to access U.S.-controlled technology under EAR/ITAR without sponsorship for an export license.

Nice to have

  • Multimodal perception using EO/IR with radar, LiDAR, or RF at the feature or decision level.
  • Physics-aware imaging, radiometric correction, NUC/FFC, atmospheric-effects modeling, and synthetic-data simulation.
  • MLOps and data infrastructure experience with SQL, Parquet, dataset versioning, CI-based validation, and scalable multi-GPU training.

Culture & Benefits

  • On-site work within a rapidly growing defense technology startup of approximately 350 employees.
  • Healthcare, dental, and vision insurance, retirement savings, and paid time off.
  • Equity or stock options and performance-based incentives.
  • Funding for continuing education, training, and professional development.
  • Compensation tiers designed to be competitive, equitable, and transparent.

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 →