Machine Learning Feature Store Feature Registration and Acquisition API Backend Development

About This Role

This backend development role focuses on building the registration and acquisition APIs for a Feature Store that centralizes the features used in machine learning model training and inference. It suits a developer interested in tackling the unique technical challenges of Feature Stores, such as consistent feature delivery across training and real-time inference, and feature versioning.

Responsibilities

• Design and implement feature registration and acquisition APIs using FastAPI
• Design feature metadata tables using PostgreSQL
• Implement mechanisms to provide consistent features during training (batch) and inference (real-time)
• Implement feature version control and lineage tracking functions

Requirements

• 3+ years of development experience using Python
• 1+ year of practical experience with FastAPI or similar frameworks
• 1+ year of experience in developing and operating machine learning systems

Nice to have:

• 1+ instance of experience using Feature Stores such as Feast
• 1+ instance of experience using real-time platforms such as Kafka or Redis
• 1+ instance of practical experience with MLOps-related tools

Benefits

• Social insurance fully covered
• Remote work system
• Book purchase system
• Training and study group subsidies
• Various award systems
• Flextime system

Important Details

• Schedule: generally flexible
• Contract: contractor (業務委託)
• Holidays: 125+ days per year, full two-day weekends (Saturdays, Sundays, and public holidays), summer holidays, year-end and New Year holidays, condolence leave, paid leave

Application Process

• 1 interview

Other Conditions

• Smoking measures: principle of non-smoking at the workplace (divided smoking areas)

About the Company

This company develops products that are central to MLOps, allowing team members to acquire practical machine learning infrastructure design skills.