Preferred Networks, Inc.
- Job Type
- Machine Learning, Optimization, and Data Science Engineer
We are recruiting engineers in the fields of machine learning, optimization, and data science year-round.
We collaborate with client companies across diverse industrial sectors to solve challenges and realize PFN's vision of "making the real world computable" through research and development and implementation using our own supercomputers. We aim to provide high-value solutions that only PFN can offer, by learning specialized knowledge in each business domain and utilizing a broad range of computer science knowledge and implementation skills. We seek individuals who can actively communicate with people from diverse backgrounds, both inside and outside the company, and who can continuously learn new knowledge in line with PFN's "Learn or Die" values, thereby creating new value.
The following are examples of work; however, actual work is not limited to these. The projects and tasks you will actually be responsible for after joining the company will be determined based on your specialized knowledge and experience.
**Examples of Work**
As an engineer, you will manage and promote projects for problem-solving, including new projects.
- Launching new projects, conducting customer interviews, defining requirements, engineering, and reporting.
- Observation and analysis of data related to existing equipment and processes, as well as their operation and management.
- Defining tasks suitable for resolution using machine learning techniques and satisfying customer needs.
- Establishing and implementing evaluation methods for machine learning model operation and application behavior based on actual equipment, data, or simulations, and conducting technical verification using these methods.
- Explaining machine learning model operation and control behavior to customers and users, and developing, implementing, and verifying technical methods to enhance explainability (such as visualization).
- Providing technical advice to engineers at client companies.
We provide problem-solving services such as time series forecasting, mathematical optimization, and inverse problem analysis, depending on the problem domain and the client's business domain.
- Construction and verification of machine learning models for prediction or anomaly detection of time-series data related to target equipment and processes.
- Proposal, implementation, and experimentation of solutions to real-world combinatorial optimization problems from a multifaceted perspective including heuristics, mathematical programming, and machine learning.
- Inverse problem analysis and estimation of the internal state and structure of measured objects using real-world remote sensing data.
- Domain adaptation of simulation data and real-world data.
We develop technologies to control large-scale production facilities such as chemical plants in operation, as well as industrial machinery and robotic equipment, to enable safer and more optimal operation.
- Construction of machine learning models that enable optimal control while guaranteeing safety.
- Construction of machine learning models that detect and predict equipment anomalies using time-series operating data.
We develop technologies for efficient manufacturing, utilizing simulations and other methods, for manufacturing processes and equipment.
- Formulating problems using 2D/3D particle simulations and applying them to machine learning
- Development of continuous and discrete optimization algorithms for exploring optimal manufacturing processes
- Construction and operation of a system that performs optimization processing using developed algorithms and machine learning models in conjunction with our own cluster
We propose data analysis and problem-solving solutions that lead to improved profitability and customer satisfaction in the retail industry. We develop methods that contribute to the restructuring of operations and improvement of operational quality in the retail industry.
- Problem setting and hypothesis building through observation of store operations and discussions with stakeholders
- Leading the construction of data collection and analysis infrastructure in collaboration with other internal teams and client companies
- Development of image and POS data analysis and optimization methods for the purpose of sales promotion, avoiding lost opportunities, and inventory optimization
- Verification of the effectiveness of development methods in collaboration with client companies
