MI-6 Co., Ltd.
- Job Type
- [PD] Machine Learning Engineer
This role involves developing, improving, and implementing machine learning algorithms (Bayesian optimization, modeling, LLM utilization, etc.) integrated into our products.
[Mission]
MI-6's mission is to "promote 'materials informatics' and innovate research and development." We are developing services aimed at realizing research and development experiences where digital and physical technologies are seamlessly integrated. We contribute to materials and manufacturing in Japan and around the world, participating in the creation of a richer society.
Currently, miHub is continuously used by many major materials and manufacturing companies, and is transitioning from a "stage of increasing adoption" to a "stage of deepening utilization." Simultaneously, we are developing a next-generation product suite that strengthens integration with advanced informatics and lab automation.
Machine learning engineers are responsible for developing, improving, and implementing machine learning algorithms (Bayesian optimization, modeling, LLM utilization, etc.) integrated into MI-6 products, including miHub.
This role involves bridging research and implementation to enhance product value through activities such as speeding up and improving the accuracy of existing algorithms, introducing new algorithms, and optimizing the product experience using LLM (Limited Licensing Module).
Furthermore, those with an understanding of web application development can design algorithms and products as a unified whole, providing value from a full-stack perspective.
■ Role
1. Algorithm Development and Acceleration (Bayesian Optimization and Model Development)
- Development of new features for Bayesian optimization algorithms
- Algorithm improvement for improved search efficiency (higher accuracy, robustness)
- Algorithm and infrastructure improvement for faster computation
- Modeling centered on Python (PyTorch / BoTorch)
2. Product Quality and Accuracy Improvement
- Benchmark design and evaluation platform construction for improving algorithm accuracy
- Construction of automated evaluation and testing mechanisms to ensure model quality
- Integration into product interfaces (API / Serving)
3. Collaboration with Other Teams
- Algorithm specification formulation in collaboration with researchers and PdMs
- Internal sharing of developed algorithms (documentation and design explanations)
- Product integration and bridging to UI/UX in collaboration with Web engineers
【Career Growth】
You can pursue leadership positions in both technical contribution and organizational development.
Furthermore, by being involved in product strategy and business operations, you can also consider a career path as a management executive, leveraging your organizational management skills and technical expertise.
【Job Appeal】
・You can challenge yourself with complex ML engineering unique to the research and development field, combining Bayesian optimization, model search, LLM, algorithm acceleration, etc.
・A full-stack environment where you can develop across the boundaries of ML/algorithms and web products on a monorepo architecture
・You will acquire comprehensive skills in algorithms, infrastructure, and optimization
・Many former materials researchers and data scientists are employed in the company, allowing you to develop while deeply understanding user problems
・You will gain experience in creating an R&D platform aiming for industry standards
【Technology Stack】
・Frontend: TypeScript, React
・API: Ruby on Rails, Python
・ML: Python (PyTorch, scikit-learn, etc.), Rust
・Infrastructure: AWS (ECS Fargate, Aurora, etc.), Terraform, GCP (BigQuery), CircleCI, GitHub Actions, Datadog
・Other: GitHub, Slack, Jira
【Selection Flow (Expected)】
Document Screening
→ Multiple Interviews (approximately 2-3 times)
→ Reference check/background check
→ Final interview
