Pairing with Data Scientists, develop machine learning software product including but not limited to explore large data set, try out new algorithms, feature engineering, test and evaluate model output, deploy the solution for production usage and also scale out to the comprehensive fashion network of a retail company.
Design, develop and maintain the large-scale data infrastructure required for the machine learning projects
Leverage on understanding of software architecture and software design patterns to write scalable, maintainable, well-designed and future-proof code
Develop tool and framework to address common needs in machine learning project like model traceability, feature reuse, A/B test, etc.
Work in cross-functional agile team of highly skilled engineers, data scientists, business stakeholders to build the AI ecosystem within organization.
Have a degree within software engineering or similar field (e.g. computer science and programming) and a strong will to continuously develop your engineering skillset
Are a hands-on person, love coding as much as breathing and would like to deploy software engineering practice into machine learning projects.
Have at least 5 plus years of working experience in developing machine learning products into production
Are experienced in one of following programming languages: Python, Scala or Java. Experience in Spark and distributed computing is seen as a merit.
Have experience with handling high volume heterogeneous data (both batch and stream) and good understanding about data storage and data structure would be a big plus
Have experience from working with cloud solutions
Have experience in agile environment, team collaboration, data-driven development, reliable and responsible experimentation
Meriting not required: Kubernetes, Airflow or other scheduling tool
Required language skills:
Work with a pricing engine for online markdown (sale) prices. The team delivers price recommendations for online sale periods, which happen four times a year. The core components of the product is a demand prediction for upcoming volume of sale depending on discount level and optimization to find the right price that maximizes the net sales.
The product works in a setting with Azure cloud, Kubernetes, Airflow and is mainly programmed in Python and Spark (utilized through Databricks).