We’re hiring a computer vision engineer to work with our small, full-stack team to build products to test, monitor, and improve the reliability of machine learning systems. We work with customers that deploy computer vision systems across a wide range of high-risk domains, and they rely on our products to ensure that their systems are reliable and interpretable.
Responsibilities
As a computer vision engineer, you’ll be responsible for developing state-of-the-art methods to test the robustness and interpretability of computer vision models. You’ll also be responsible for researching different computer vision problem domains and working with customers in those domains (e.g. medical imaging, autonomous driving, manufacturing).
Our backend is primarily in Python, and we use many Python ML libraries (PyTorch, Tensorflow, etc.). Some other (non-ML) technologies we use are: Docker, PostgreSQL, Redis, RQ, Gunicorn, Flask, TensorFlow Extended, Pytest.
Requirements
Our company, technology, and market are new and changing rapidly, so a large part of the job is to adapt and learn new things. We value learning fast over pre-existing knowledge (“slope is more important than y-intercept”).
You should have at least 1 year of computer vision software engineering experience. We don’t require past experience with specific programming languages or frameworks.
We prioritize candidates in Tokyo, but may consider remote for exceptional candidates.
Nice to haves
These aren’t required, but be sure to mention them in your application if you have them.
You can work full-stack, you’ve worked at a startup before, you have experience with production machine learning, or you have research experience in machine learning.
Compensation
As an early employee, you’ll own a significant equity stake in addition to your base salary.