This is a Data Science role with a much stronger production engineering expectation than a typical modelling-only position.
We are looking for someone with strong experience in:
•
Recommendation systems, ranking or personalisation.
•
Python in production Data Science environments.
•
Large-scale customer, product or behavioural datasets.
•
Object-oriented programming.
•
Clean code and software design principles.
•
Modular, reusable and testable ML code.
•
CI/CD practices for Data Science or ML workloads.
•
Model evaluation and experimentation.
•
Deploying models into production.
•
Monitoring and maintaining production models.
Relevant recommendation experience could include:
•
Content-based recommendation;
•
Experimentation and incremental impact measurement.
Experience with MLflow, Delta Lake, Databricks Workflows and model lifecycle management would be highly valuable.
The strategic direction already exists.
The successful person needs to be comfortable coming into an established environment, understanding the current state quickly, and executing and improving the agreed approach.
You will likely be someone who has spent several years as a strong hands-on Data Scientist and has gradually taken on more responsibility for how your models are engineered, deployed and operated.
You should be equally comfortable:
•
Discussing recommendation methodology;
•
Reviewing model performance;
•
Designing clean class structures;
•
Working through production issues;
•
Explaining technical trade-offs to stakeholders.
You will need to be pragmatic.
We don’t need someone who immediately wants to redesign everything. They need someone who can understand the existing plan, challenge it where necessary, and then drive it through to a high-quality implementation.
You will also be expected to:
•
Build credibility quickly with senior client Data Science stakeholders;
•
Operate independently within the client team;
•
Communicate progress, risks and technical decisions clearly;
•
Work collaboratively rather than positioning yourself as an external reviewer;
•
Balance client delivery priorities with good engineering practice;
•
Leave the recommendation capability in a stronger and more maintainable state than you found it.