Employer: MHA Management Holdings
Industry: Financial Services, Long-Term Funeral Insurance, Applied AI Engineering & Data Governance
Department: Data & AI Team
Location: Bedfordview, Johannesburg, Gauteng, South Africa (On-Site)
Position Type: Full-Time Graduate Internship Programme (Clear Career Growth Path to Data & AI Engineer)
Target Candidates: Recent Graduates / Final-Year Students in Data Science, Computer Science, IT, Stats, Maths, or Engineering
EE Status: Employment Equity / Affirmative Action (EE/AA) & Student Open
Compensation: Market-Related Corporate Internship Allowance
Posted Date: 24 October 2026
About MHA Management Holdings & The AI Internship
MHA Management Holdings is an established leader in the South African long-term insurance market with 30 years of operational experience, specializing in innovative funeral insurance administration across the trade union sector.
MHA’s Data & AI team operates a production-first, highly governed Microsoft data estate featuring SQL Server, Azure SQL, Microsoft Fabric lakehouses, Power BI/SSRS reporting, Python pipelines, and Azure DevOps CI/CD pipelines.
MHA is recruiting a curious, detail-oriented Graduate: AI Intern to join their applied-AI team in Bedfordview, Johannesburg. Designed as an engineering-and-governance training pathway (not a research post), this internship provides hands-on exposure to preparing datasets, running in-house governed AI platforms, testing summarization/classification/RAG search use cases, writing Python scripts, and applying POPIA data minimization standards. High-performing interns serve as the direct talent pipeline for permanent Data & AI Engineer roles.
Key Responsibilities & Applied AI Workflows
The AI Intern will execute nine core operational and engineering deliverables under direct mentorship:
1. Data Preparation & Engineering Pipelines
- Dataset Readiness: Prepare, clean, validate, and document datasets consumed by internal AI, retrieval, and search workloads.
- Python Automation: Write clean Python scripts and Jupyter notebooks to automate repetitive data preparation, extraction, and testing routines.
2. In-House Governed AI Platform Support
- Platform Operations: Assist in managing the governed AI platform—handling content ingestion, document indexing, configuration tweaks, and system health checks under supervision.
- Use-Case Testing: Test and evaluate AI use cases (summarization, classification, vector search, data extraction, and workflow automation) against strict accuracy criteria.
3. Governance, Evaluation & POPIA Compliance
- Evaluation Logs: Keep detailed records of evaluation evidence, tracking raw AI outputs against human reviewer modifications.
- Data Anonymization: Apply strict data-minimization and anonymization rules (POPIA/FSP environment), ensuring zero personal credentials or unapproved data enter external tools.
- Documentation: Maintain data dictionaries, prompt engineering configuration notes, use-case write-ups, and testing logs.
Candidate Profile & Technical Entry Requirements
To qualify for consideration, applicants must meet the following mandatory technical and personal criteria:
- Core Technical Skills (MANDATORY DAY-ONE BASICS):
- Foundational Python: Must be able to read, run, and modify Python scripts (gained through coursework, projects, or self-teaching) (Mandatory).
- SQL & Relational Data: Foundational understanding of SQL queries and relational database concepts (Mandatory).
- Educational Qualifications:
- Qualification in progress or recently completed in Data Science, Computer Science, Information Systems, Statistics, Mathematics, Engineering, or a related quantitative field (Advantageous).
- AI Mindset & Exposure:
- Practical curiosity about modern AI tools paired with healthy skepticism regarding model hallucinations and outputs (Mandatory).
- Advantageous: Exposure to APIs, notebooks, vector search / retrieval-augmented generation (RAG) concepts, or LLM tooling.
- Advantageous: Project portfolio, capstone, or hackathon evidence demonstrating data/AI problem solving.
- Personal Attributes:
- High level of integrity when handling sensitive POPIA/financial data.
- Growth mindset, composed under pressure, with strong written and verbal communication skills.
Application Guidelines & Online Submission
- Prepare Document Package:
- Detailed Curriculum Vitae (CV) highlighting Python, SQL, and any AI/data projects (GitHub or capstone links encouraged).
- Certified Copy of South African ID Document or Passport.
- Certified Copies of Tertiary Qualification Certificates or Official Academic Transcripts.
- Online Portal: Submit your application directly via the official MHA Management Holdings Careers Portal.
- Closing Date: Applications are processed on a rolling basis; early submission is strongly recommended.
Sample Selection & Applied AI Interview Questions
Python, SQL & Responsible AI Governance
- Question (AI Skepticism & Testing): “Can you give an example of a project or scenario where an AI tool (like an LLM) generated an incorrect or hallucinated output, and how did you detect and fix the error?”
- Question (POPIA & Data Minimization): “When preparing customer data for an internal search or indexing pipeline, what exact steps do you take to sanitize PII (Personally Identifiable Information) before processing?”
How to Apply
Apply online directly via the official MHA Management Holdings Careers Portal:
- Employer: MHA Management Holdings
- Location: Bedfordview, Johannesburg, Gauteng
- Position: Graduate: AI Intern (Applied AI & Data Engineering)
- Target Group: Graduates/Students in Data Science, CS, IT, Stats, Maths, Engineering