[Japan] Machine Learning Engineer - Kawasaki, Kanagawa, Japan
Deploy neural networks on the SAKURA AI Chip and proprietary edge AI platform using PyTorch and TensorFlow.
About the role
This role is part of EdgeCortix's SAKURA AI Chip team, focused on deploying neural networks on proprietary edge AI hardware and software stacks. You will work on neural network implementation, quantization, and deployment using PyTorch and TensorFlow.
What you'll do
- Work closely with compiler and hardware engineers to support deployment of neural networks on EdgeCortix's proprietary hardware and software stack. - Guide the deployment of custom and off-the-shelf models to meet accuracy and performance targets for customer AI systems. - Grow and maintain a collection of ready-to-use models for perception tasks. - Implement models from scratch, train, quantize, prune, and deploy models on EdgeCortix hardware while conducting accuracy studies. - Review academic literature, work with public and proprietary vision datasets, and collaborate continuously with hardware and compiler teams.
What you'll need
- Experience in writing Python code. - Experience using TensorFlow and PyTorch ML frameworks. - Strong knowledge of Deep Neural Networks. - Strong understanding of network implementation, training, and accuracy evaluation techniques. - Strong debugging and analysis skills for root-causing complex issues. - Experience with Git and GitHub/GitLab development flows.
Nice to have
- Pursuing a Master in Computer Science, Data Science, or similar. - Strong object-oriented design and development skills. - Knowledge of SOTA neural network architectures for perception tasks (such as object detection, semantic segmentation, monocular depth estimation) and vision datasets (COCO, KITTI). - Experience in usage of deep learning compiler frameworks such as TVM, Glow, or XLA. - Experience in model deployment and development of ML applications on an embedded device. - Experience with network quantization and pruning techniques. - Experience with customizing network implementations and hyperparameter tuning. - Knowledge of RNN/LSTMs, Transformer architectures, point cloud processing, sequence networks (NLP), and graph processing networks (GNN). - Knowledge of Neural Architecture Search (NAS) techniques. - Knowledge of Reinforcement Learning.
Skills: neural networks, PyTorch, TensorFlow, quantization, edge AI platform
This role has been open 124 days — well above the 89-day median for AI/ML Hardware Engineering roles.
AI/ML Hardware Engineering · AI ML Hardware
|
Open roles in category
507
|
Median days open
89 d
|
Median salary
$236k
|
See the full market breakdown ▾Category comparison, skills in demand, and who else is hiring
| Metric | EdgeCortix | All employers we track in this specialty (507 roles · 76 employers) |
|---|---|---|
| Open roles in this specialty | 1 | 507 |
| Open roles in the wider Software, Firmware & Systems family | 4 | 6292 · 146 employers |
| Median days open | 124 d | 89 d (+35 d vs this employer) |
| Median salary (USD postings) | — | $236k |
Skills observed across this category: neural networks, PyTorch, TensorFlow, quantization, edge AI platform
Who's hiring in this category
- Qualcomm · 104 open roles · median 118 d
- NVIDIA · 88 open roles · median 60 d
- AMD · 32 open roles · median 110 d
- Micron Technology · 27 open roles · median 62 d
- Mobileye · 20 open roles · median 88 d
- Analog Devices · 18 open roles · median 45 d
How we counted: 507 open AI/ML Hardware Engineering (AI ML Hardware) roles from 76 employers tracked in the SemiconductorJobs index, counted 27 Oct 2026. Specialty figures count only roles carrying this exact specialty label, so an employer's related work in neighbouring specialties is not included there — it is counted in the wider Software, Firmware & Systems family row. Figures refresh nightly.