Data Scientist

Cloud23; Johannesburg; On-site; ZAR 900,000 - 1,500,000; Full time

Job summary

Cloud23 in Johannesburg is seeking an experienced data scientist to analyze complex enterprise data, frame business questions as analytical problems, and build statistical and machine-learning models that inform decision making.

You will work with domain experts in financial services, validate findings, communicate uncertainty, and hand models to ML engineers for production, ensuring reproducible results.

Qualifications

• Solid grounding in statistics and experimental design.
• Ability to translate questions into defensible analyses.
• Experience turning analysis into actionable insights.

Responsibilities

• Frame business problems as analytical ones, and be honest when the data cannot answer the question asked.
• Build, evaluate and iterate statistical and machine-learning models against real client data.
• Work directly with domain experts in banking, insurance and investment management to validate what the numbers mean.
• Communicate findings to people who will make decisions on them — clearly, with the uncertainty stated.
• Hand models to ML engineers in a state they can productionise, with evaluation and assumptions documented.

Skills

Statistics

Experimental design

Data analysis

Python

SQL

Communication

Uncertainty handling

Job description

Analyse complex enterprise data, find the patterns that matter and build the statistical models behind decisions banks actually act on.

What you’ll do

• Frame business problems as analytical ones, and be honest when the data cannot answer the question asked.
• Build, evaluate and iterate statistical and machine-learning models against real client data.
• Work directly with domain experts in banking, insurance and investment management to validate what the numbers mean.
• Communicate findings to people who will make decisions on them — clearly, with the uncertainty stated.
• Hand models to ML engineers in a state they can productionise, with evaluation and assumptions documented.

What we’re looking for

• Strong grounding in statistics, experimental design and the limits of both.
• Fluency in Python and SQL; comfort with the modern analysis stack.
• Experience taking an analysis from question to defensible conclusion, not just to a notebook.
• The judgement to say when a simpler model is the right answer.
• Financial-services or other regulated-domain experience is an advantage, not a requirement.