Group Senior AI Engineer at Salt South Africa

What You'll Be Doing


This is the senior-level version of an AI Engineer: same end‑to‑end ownership, but with more leadership, more influence, and bigger, messier, more rewarding problems.


You'll:


Take full technical ownership of AI systems - from prototype to production reliability.
Make architectural calls that balance speed, stability, and long-term maintainability.
Convert loosely defined business questions into clear AI solutions with measurable impact.
Build everything from classical ML models to LLM-powered services (RAG, fine-tuning, adapters, you name it).
Deeply integrate your models into apps, CRM platforms, workflows, digital journeys, and internal tools.
Lead experiments, run A/B tests, validate hypotheses, and guide stakeholders through trade‑offs.
Set up monitoring, drift checks, retraining flows, observability, and proper MLOps hygiene.
Work hands-on with data, modelling, pipelines, APIs, orchestration - the full engineering spectrum.
Mentor juniors, shape engineering standards, and help create reusable blueprints for future AI builds.
Act as a champion for safe, transparent, responsible AI use.
If you like being the person who steps into complexity, simplifies it, owns it, and ships something solid - good news: you've found your people.


Who You Are
You don't need to match every bullet, but you're likely to thrive if you're someone who:


Has roughly 8+ years working in ML, data science, or software engineering with real AI ownership.
Has shipped multiple AI/ML systems into production - not just POCs.
Has led at least one meaningful end‑to‑end AI initiative (technical direction + stakeholder alignment).
Is comfortable with Python + SQL and enjoys writing clean, production-level code.
Knows ML frameworks (scikit‑learn, XGBoost, PyTorch, TensorFlow, LightGBM etc.).
Has experience with LLMs and generative AI (APIs, RAG, evaluation, tuning).
Understands data engineering basics and can work with pipelines, warehouses, and orchestration tools.
Is familiar with the cloud (Azure, AWS, or GCP all welcome).
Navigates ambiguity well - shaping scope instead of waiting for it.
Communicates clearly, especially with non‑technical stakeholders.
Loves mentoring, pair programming, reviewing code, and raising team standards.
Understands privacy, safety, and the need for responsible AI.
Values outcomes over grand ideas.