IGrow Wealth Investments Group of Companies, proudly positioned as South Africa's #1 Investment company, is searching for an ambitious and practical Senior AI Engineer to join our growing team in Durbanville, Cape Town.
This is a hands-on engineering role for someone who is already a strong software engineer and understands how modern AI changes the way applications, workflows and business processes are designed and built.
You will design and build AI-first applications, agents and automated workflows that integrate with real business systems and data.
You will work across backend engineering, AI systems, data architecture and agentic automation, including building systems where AI agents can reason, use tools, access organisational knowledge, execute workflows and collaborate with people.
This is not primarily a data-science or prompt-engineering role. We are looking for a strong engineer who can build production software around modern AI.
You should be able to move between conventional software engineering and AI engineering rather than treating them as separate disciplines, understand the important principle of modern AI development.
"AI is not just a chatbot. It can be an active participant in software systems finding information, using tools, making decisions within defined boundaries, coordinating workflows and working with humans to achieve outcomes. You will be expected to use AI extensively in your own engineering workflow. We want engineers who understand how to direct coding agents, review what they produce, test it, improve it and remain accountable for the resulting software."
You should be capable of taking a business requirement, helping shape the architecture, making sound technical decisions and delivering robust production software with limited supervision.
Minimum requirements (Advantageous):
Advanced commercial Node.js experience.
Advanced TypeScript.
Strong backend software engineering and API design.
Strong PostgreSQL knowledge.
Solid understanding of relational data modelling and database performance.
Practical understanding of LLMs and modern AI application development.
Understanding of embeddings, semantic similarity and vector search.
Experience with or strong understanding of vector databases / pgvector.
Understanding of graph databases and knowledge graphs.
Understanding of agentic AI, tool calling and multi-step AI workflows.
Experience building integrations between different systems and APIs.
Comfortable with asynchronous processing, queues, scheduled jobs and event-driven architectures.
Strong Git and modern software-development practices.
Experience using AI coding agents/coding assistants as part of serious software development, not simply autocomplete.
AWS, Docker, ECS, S3, Redis, Bedrock, OpenAI, Anthropic, Mastra, MCP, RAG architectures, pgvector, Neo4j or other graph databases.
OAuth/OIDC, event-driven architectures, workflow engines and infrastructure as code.
Duties and responsibilities include, but not limited to:
Build production AI applications and agentic systems using Node.js and TypeScript.
Design backend services, APIs, integrations and event-driven systems.
Build agentic workflows where AI agents use tools, data and other agents to complete real business tasks.
Work with LLMs from multiple providers and understand model selection, context management and tool/function calling.
Design and implement AI memory and organisational knowledge systems.
Work extensively with PostgreSQL, including schema design, complex queries, performance and production data architecture.
Implement embeddings, semantic search, RAG and vector retrieval, including PostgreSQL/pgvector and dedicated vector databases where appropriate.
Understand and work with graph data and graph databases, including relationships between people, systems, processes, documents and organisational knowledge.
Build integrations with internal and external systems through APIs, MCP servers and other tool interfaces.
Design reliable long-running AI processes involving scheduled events, human approvals and asynchronous execution.
Use modern AI coding agents as a core part of the development process while maintaining engineering quality, testing and code review.
Help establish engineering patterns, reusable components and technical standards for AI-first development.