MLOps Engineer

【About the company ...】

The company is a software company whose corporate message is "Speed up your Business.

They accelerate their customers' businesses in various fields such as medical, manufacturing, finance, and entertainment by parallelizing and optimizing software to efficiently utilize multi-core processors and accelerating applications using new memory technology that realizes power saving and high-speed IO, thereby achieving green IT.

【 Job Description 】

【Overview】
They are developing machine learning infrastructures for efficient development of high-performance machine learning models.
Specifically, they are researching and developing lightweight methods such as quantization and branch pruning of deep learning models, domain-specific compilers that automatically speed up the inference process of models, and HW aware NAS infrastructure that can be recognized efficiently on the target hardware, and building MLOps using these methods.

【Responsibilities】
・Development of deep learning compilers
・Research and development of lightweight methods for deep learning models such as quantization and branch pruning
・Development of HW aware NAS
・Construction and operation of MLOps using the above

【Rewarding】
・Challenge new technological innovations in the fast-growing AI market
・Contribute to the development of well-known OSS
・Experience building systems using public cloud services
・Surrounded by top-notch engineers, improve your skills through daily work and discussions

【Development Environment】
Development environment: C, C++, Python3, Bash
Framework：TensorFlow
Other development environment: Linux, Debian
Development Support Tools: JIRA, Git, GitLab, Apache Airflow
Development Methodology: Agile, Scrum, Global team (multinational members)
Development content type: B2B, machine learning, AI, mobility-related (self-driving, traffic-related)
Cloud platform: Amazon Web Service
Infrastructure management: Docker, Terraform, Ansible, Kubernetes, Amazon ECS, Amazon Elastic Kubernetes Service
AI and data analysis: Apache Beam, Keras, PyTorch, pandas, Jupyter notebook, matplotlib, NumPy

--------------------------------------------------

【 Working time 】

Discretionary Work System

【 Welfare 】

・Full social insurance
・Commuting allowance
・Child allowance
・Recreation Assistance
・Training/Self Development AllowanceResident allowance of 30,000 yen/month
・Travel and accommodation expenses for company trips
・Company expenses for social gatherings
・Maintenance fee for club activities (e.g., ski club: 60,000 yen per person per year; other clubs include golf, futsal, and board game clubs)

【 Holiday 】

• 2 days a week (Saturday and Sunday)
• National Holidays
• Summer vacation
• New Year holiday
• Annual Paid leave, etc.

[Required]
・Experience in development using C/C++ and Python
・Basic knowledge of deep learning

[Preferred]
・Experience using or developing deep learning frameworks and deep learning model formats such as TensorFlow, PyTorch, ONNX, etc.
・Experience developing domain-specific compilers such as TVM, Halide, MLIR, etc.
・Experience in software acceleration of deep learning inference processes
・Knowledge or development experience with deep learning accelerators
・Development experience using cloud infrastructure such as Amazon Web Service or Azure
・Experience in building and operating MLOps