Description

• Homechoice is a leading South African homeware retailer. For almost 40 years we’ve helped our customers create beautiful homes they love with an innovative range of quality products they can afford.
• The purpose of this role is to create and maintain statistical and machine learning models in production, and to distribute the model output to various teams within the business. The Data Science team maintains predictive machine-learning models in production, whose outputs guide decision-making in Marketing, Credit Risk and the Call Centre.

What you will love doing in this role

• Maintaining and monitoring of existing predictive models in production
• Ensures that existing model performance does not degrade over time.
• Picks up model errors early (prior to model outputs being used by the business)
• Writes code to monitor the historic performance of existing models.
• Writes code to monitor the integrity of the mode input data.
• Reporting (accurate, informative and scientific)
• Delivers consistent reports in excel, which do not contain any errors.
• Objectively quantifies the uncertainty surrounding estimates (e.g. confidence intervals)
• Presents results clearly, in a way that clearly guides business decision-making.
• Adherence and contribution to best practice (statistical and software)
• Knows and understands the existing team best practice specifications.
• Implements best practice in their daily work.
• Contributes to the team best practice specifications.
• Designing new predictive models
• Creates new predictive models from scratch (the full pipeline)
• Ensures that new models adhere to the standard team model structure.
• Creates model documentation allowing zero-day model handover.
• Maintaining and contributing to the data science tool kit (code base)
• Effectively uses the tools in the existing team code base
• Recommends tools that can be added to the existing team code base, and ones that can be improved
• Adds new code to the existing team code base (with sufficient testing)
• Continuous Automation of Regular Processes
• Writes code that reduces the time taken to run regular processes, without reducing the output quality.
• Ensures that automated processes generate sufficient logging that the team maintains trust in the output.

Requirements
What you’ll need to do this role

• Undergraduate degree containing mathematics, statistics or computer science.
• Evidence of ongoing personal development in data science, software development, statistics or machine-learning (online courses, personal blog etc.
• Previous data science, machine learning or statistical role.
• ML-Ops (running models in production)
• Previous experience in Retail and/or Credit sector.
• Advanced Python & SQL skills.
• High proficiency in MS Excel.
• Statistics and probability theory.

Behaviors we love

• Wow my customer
• Walk in my customers’ shoes
• Deliver on my promises
• Deliver insight-led solutions my customers need
• Treat the business as my own
• Take accountability
• Be curious, creative & explore opportunities
• Do it right & at the right time
• Play as a team
• Be helpful
• Be inclusive
• Find the fun