SENIOR DATA SCIENTIST – Build Models That Actually Make It Into Production – CAPE TOWN (Hybrid), R1.1m – R1.5m

This is an excellent opportunity for a SENIOR DATA SCIENTIST who wants to do more than build good models.

Based in CAPE TOWN, hybrid, this role offers a salary of R1.1m – R1.5m.

This SENIOR DATA SCIENTIST role is with one of South Africa’s most ambitious technology-led FinTech businesses, where data sits at the heart of how products are built, customers are understood, risk is managed and decisions are made.

The opportunity is particularly attractive if you enjoy the full journey from identifying the problem, analysing the data and building the model, through to getting that model into production and measuring whether it is actually delivering value.

THE COMPANY:

This is a high-growth FinTech business using technology and data to solve one of the biggest challenges facing South African SMEs: access to fast, intelligent and frictionless financial services.

This is not a traditional financial-services organisation trying to retrofit technology around an old operating model. Data, software and intelligent decision-making are central to the business.

They have built a sophisticated digital platform serving South African SMEs and are continuing to invest heavily in their Data capability as the business scales.

For a strong Data Scientist, this creates an unusually interesting environment: the work is commercially important, the datasets are real, the problems are complex and the models you build have a direct impact on business decisions.

THE ROLE:

As SENIOR DATA SCIENTIST, you’ll work across product, operations, finance, marketing and credit to turn business questions into analytical problems and data into better decisions.

You’ll be responsible for more than experimentation.

You’ll help frame the problem, explore the data, build and validate models, and take responsibility for getting those models into production and monitoring how they perform once they are there.

You’ll work on areas including:

Predictive modelling, classification and regression, forecasting, segmentation, experimentation and A/B testing, feature engineering, model evaluation, productionising ML models, model monitoring and drift detection, retraining and model improvement, large-scale data analysis, reusable analytics and reporting.

You’ll work closely with Data Engineers and Analytics Engineers, rather than being expected to personally build every layer of the underlying ML platform.

You’ll also help raise the standard of Data Science across the team through good code, reproducibility, testing, documentation and technical guidance.

THE TECH:

The broader environment includes:

Python, SQL, Azure, Snowflake, dbt, Airflow, Airbyte, Event Grid, CI/CD and MLOps tooling.

Exposure to technologies such as Docker, MLflow, workflow orchestration or Infrastructure as Code would be useful, but you are not expected to arrive having worked with every technology in the stack.

REQUIRED SKILLS:

You’ll ideally have around 3–5 years’ experience in Data Science, advanced analytics or a closely related field, together with a strong quantitative foundation.

You’ll need:

Strong Python experience for analysis and modelling.

Strong SQL.

Experience building and evaluating ML models on real-world data.

Solid statistical and analytical capability.

Experience taking models beyond experimentation and into production.

Exposure to monitoring model performance, drift and data quality.

Good software-engineering discipline, including Git, code review and testing.

The ability to communicate clearly with both technical and non-technical stakeholders.

I’m particularly interested in candidates who can show that their work has influenced real business decisions rather than simply producing technically interesting models.

Experience in FinTech, banking, lending, credit or financial services would be highly relevant, but is not essential.

WHAT WILL MAKE YOU STAND OUT

You may have experience with some of the following: MLflow, DVC or similar MLOps tooling, dbt, CI/CD, cloud data platforms, Snowflake, Docker, Airflow or other workflow orchestration tools.

Most importantly, you’ll be someone who understands both sides of the job:

How do I build a model that works?

And:

How do I make sure it continues to work once it is being used in the real world?

If you are a Data Scientist who wants your work to move beyond the notebook and into meaningful production use, this is the kind of opportunity that can take your career forward.