Agentic AI Engineer

About the role

Ozow, a fintech company in South Africa focused on digital payments, is looking for an Agentic AI Engineer in Cape Town. The role develops and deploys AI agents that automate business processes, support customers, strengthen fraud and risk management and improve engineering workflows. The engineer works closely with experienced engineers and must ship at least one agent feature into production within the first six months, with robust evaluation tests.

Key responsibilities

• Work with product, operations and risk management teams to understand the business processes the agents will support.
• Propose new automation use cases and identify manual processes suitable for automation.
• Build and maintain agent components: tool calling, Retrieval Augmented Generation (RAG), prompt design and multi-step task orchestration.
• Connect agents securely to internal APIs and data sources.
• Write clean, tested code that passes code reviews.
• Develop evaluation tests for accuracy, reliability and safety, and monitor deployed agents and failures.
• Document each agent and follow change management, release and security protocols.
• Support business users in adopting the tools and share knowledge of agent frameworks, Model Context Protocol (MCP) and new models.
• Handle personal and financial data in line with POPIA and internal policies.

Required profile

• Degree in Computer Science, Engineering or a related field, or equivalent practical experience.
• Strong software engineering foundation and interest in AI.
• Security-conscious approach to data handling.

Required skills

• Python, REST APIs and JSON.
• Hands-on experience with at least one LLM API.
• Foundational understanding of RAG, embeddings and prompt engineering.
• Git and Docker; exposure to an agent framework such as LangGraph, CrewAI or an agent SDK, or MCP.
• Desirable: AWS, C#/.NET, SQL Server or PostgreSQL, and experience in payments, fraud or financial services.