AI & Data Engineer

AI & Data Engineer We are looking for an AI and Data Engineer who can operate across the full delivery stack – from cloud infrastructure and data pipelines through to LLM application development and client-facing requirements discovery. This is a hands-on role for someone who builds things, ships fast, and is comfortable with ambiguity. You will work across multiple client engagements simultaneously, own your workstreams end to end, and contribute to a growing technical practice. JOB DESCRIPTION INCLUDES 1. Cloud Infrastructure & DevOps • Provision and manage cloud infrastructure on AWS or Azure. Deploy and maintain environments for active client engagements. Own your environments – do not wait for someone else to unblock you. • Write infrastructure as code using Terraform or equivalent. All environments are version-controlled, reproducible, and deployable without manual steps. • Containerise applications using Docker. Implement CI/CD pipelines using Azure DevOps, GitHub Actions, or equivalent. • Apply identity, access, and networking fundamentals across all environments. IAM, RBAC, SAS tokens, VPC configuration. Security is not optional. 2. Data Engineering & Integration • Design and implement data pipelines and orchestration workflows. Work across SQL, data modelling, and Python. Handle structured and unstructured data from multiple sources. • Build and integrate APIs using REST, FastAPI, or equivalent. Connect disparate client systems using appropriate integration patterns. • Work with messy, undocumented client data and make it useful. Apply data quality and transformation approaches that are pragmatic and fast rather than perfect and slow. 3. LLM & AI Application Development • Build LLM-enabled applications: prompting, RAG architectures, embeddings, and evaluation frameworks. Use open-source and commercial LLM providers as the use case demands. • Design and implement RAG systems for document-heavy environments. Handle chunking, embedding, vector store selection, and retrieval quality evaluation end to end. • Build evals and testing frameworks for LLM outputs. AI application behaviour must be measurable and reproducible. • Apply prompting best practices and LLM application patterns appropriate to the use case. Do not over-engineer. The measure of success is a working system in front of the client. • Document AI application architecture so the client team can operate it after you leave. Handover is part of the job. 4. Consulting Delivery • Take problem briefs from the Managing Partner or Technical Lead and turn them into working solutions. Not everything will be clearly defined. Asking the right questions, scoping tightly, and delivering something demonstrable is the job. • Scope a POC down to the smallest provable increment. Prioritise rapid prototyping over polished engineering. Speed of iteration matters more than perfection in the early stages. • Estimate accurately and hit short timelines. Communicate blockers early to the Technical Lead. If something is going to be late, say so before it is too late to do anything about it. • Contribute to a growing technical practice. Share what you learn, document what you build, and help make the next engagement faster than the last one. TECHNICAL REQUIREMENTS Formal qualifications are not the primary criterion. We want to see evidence of delivery. Show us what you have built. Core technical stack • Cloud: AWS or Azure — at least one deeply; compute, storage, networking, identity • Infrastructure as Code: Terraform or equivalent • Containerisation: Docker, basic orchestration • Languages: Python — data wrangling, scripting, glue code — and SQL You should also have • Data pipelines and orchestration — you have built them and they ran in production • APIs and integration patterns — REST, FastAPI, or equivalent • LLM application work — prompting, RAG, embeddings, evals. Not just experimented with. Built something. • CI/CD — Azure DevOps, GitHub Actions, or equivalent • Observability and debugging in environments you did not build WHAT WE ARE LOOKING FOR • Someone who is comfortable with ambiguity. Client requirements are often vague, data is often messy, and timelines are always short. That needs to excite you, not frustrate you. • A track record of scoping and shipping POCs and MVPs quickly. We value speed of learning and delivery over years of experience. • Someone who communicates proactively. Blockers get raised early. Nothing is hidden until the deadline. • A genuine interest in the practical application of AI – not the hype. You read the papers and you also ship the code. • Based in or able to work aligned to South African business hours. The role is remote-first with occasional on-site requirements. TECHNICAL SKILLS • Programming: Python, JavaScript frameworks — Next.js, React advantageous — SQL • Frameworks & Tools: Node.js, PyTorch, Docker, Kubernetes • Data & Integration: RESTful APIs, microservices architecture, message queues • DevOps & Version Control: Git/GitHub, CI/CD pipelines • Systems: Windows and Linux environments • Cloud Platforms: AWS, Azure, or GCP advantageous CORE COMPETENCIES • Strong problem-solving, analytical, and system-thinking capability • Ability to design scalable, maintainable, and secure software solutions • Strong communication skills with technical and non-technical stakeholders • High levels of ownership, accountability, and professional judgement