Senior AI Engineer - Production ML & Transformer Models

This is an excellent opportunity for a AI ENGINEER to Build Production ML & Transformer Models for a Leading RegTech.

100% REMOTE WORK (in SOUTH AFRICA), this AI ENGINEER role offers a salary of R1.1m - R1.2M

Unlike most AI roles involving Prompt engineering experiments, Chatbot wrappers, Internal productivity tools, and Endless proof-of-concepts - this role is different.

The company is building a regulatory intelligence platform used by financial institutions to predict complaint outcomes, surface conduct risk, and turn customer signals into actionable regulatory insight.

That means: Real data. Real decisions. Real consequences. Real production AI.

And in this role, you will own and evolve the machine learning systems at the core of the platform - from transformer fine-tuning through to production deployment at scale.

THE COMPANY:

An AI and Regulatory Technology (RegTech) company building a conduct-intelligence platform for financial services, focused on helping banks, insurers and regulated firms turn large volumes of customer feedback, complaints and operational data into structured, actionable insight.

Its technology uses machine learning and advanced analytics to identify emerging conduct risks, predict regulatory and complaint outcomes, and give compliance and leadership teams real-time visibility into how customer outcomes are performing in practice.

By connecting fragmented data across organisations into a single intelligence layer, the platform enables firms to move from reactive compliance to proactive, outcomes-driven governance and evidence fair customer treatment at scale.

THE ROLE:

This is a truly career-advancing opportunity to build real applied ML, and not just LLM wrappers.

You'll work on multi-task transformer models, prediction systems and semantic search across large regulatory corpuses.

This isn’t "call an API and hope for the best". It's proper applied ML solving complex language and decision problems.

You will own the full lifecycle (and won't be stuck in notebooks).

You'll: Guide model experimentation and fine-tuning, Build evaluation frameworks, Deploy inference services, Optimise production performance, and Monitor and iterate live systems.

The platform processes thousands of customer and regulatory signals daily: complaints, ombudsman decisions, conduct indicators and communication patterns. It's messy, nuanced, high-stakes data - which means the modelling challenges are genuinely interesting.

Essentially, you will gain: Production transformer experience, Real-world LLM deployment, ML systems running at scale, Regulatory/FS AI exposure, and End-to-end ownership.

That combination is highly valued globally.

THE TOOLS YOU’LL WORK WITH:

Python, PyTorch, HuggingFace Transformers, FastAPI, PostgreSQL & pgvector, Kubernetes / GCP, OpenAI, LangChain / LangGraph

REQUIRED:

An AI/ML engineer who likes shipping models into production.

Comfortable across modelling and engineering.

Interested in real-world impact, not just experimentation.

Keen to build systems that actually get used.

If you enjoy taking advanced ML out of notebooks and into live environments where it influences real decisions - you'll find this role compelling.

If you're currently building LLM demos or internal tools and want to step into a role where you own production AI systems end-to-end, this is a strong move.

And if you've been waiting for an opportunity to work on applied ML problems that are genuinely complex (rather than cosmetic) then please do apply.