Agentic AI Engineer

Ozow, a prominent fintech player in South Africa, is seeking an Agentic AI Engineer to join its innovative team in Cape Town. This role is at the forefront of developing intelligent agents designed to automate a wide array of business processes. If you're passionate about leveraging artificial intelligence to drive efficiency and innovation in a regulated financial environment, this opportunity could be a perfect fit.

The company is committed to revolutionising digital payments, making them more accessible, secure, and user-friendly for both consumers and businesses across the nation and beyond. By joining Ozow, you'll be part of a dynamic organisation that is shaping the future of fintech.

About the Role

This position offers a unique chance to contribute to the development and deployment of AI agents that will streamline operations, enhance customer support, bolster fraud and risk management, and optimise engineering workflows. You will collaborate closely with seasoned engineers, gaining invaluable experience in building, testing, and deploying AI solutions in a production setting. A significant aspect of this role involves ensuring the safe and responsible implementation of AI within Ozow's regulated payments ecosystem.

A key objective for new team members is to successfully ship at least one agent feature into production within their initial six months, complete with robust evaluation tests to ensure performance and reliability.

Your responsibilities will span the entire lifecycle of AI agent development and deployment. This includes:

• **Collaborating with Stakeholders:** Working diligently with product, operations, and risk management teams to gain a deep understanding of the business processes that the AI agents will support.
• **Innovation and Ideation:** Proactively identifying and proposing new automation use cases, substantiating your ideas with clear explanations of the problems they solve and the value they bring.
• **Agent Development:** Building and maintaining the core components of AI agents, encompassing tool calling capabilities, Retrieval Augmented Generation (RAG), sophisticated prompt design, and the orchestration of multi-step tasks.
• **Integration:** Securely connecting agents to Ozow's internal APIs and data sources through well-defined and robust integrations.
• **Code Quality:** Writing clean, well-tested code that adheres to team standards and passes rigorous code reviews.
• **Testing and Evaluation:** Developing comprehensive evaluation tests to meticulously check agent outputs for accuracy, reliability, and safety before deployment.
• **Production Monitoring:** Actively monitoring deployed agents, identifying and triaging any failures, and feeding these insights back into the development cycle for continuous improvement.
• **Risk Management:** Clearly articulating the failure modes of the developed agents and the mechanisms in place to detect and mitigate them.
• **Data Security and Compliance:** Handling personal and financial data with the utmost care, strictly adhering to the Protection of Personal Information Act (POPIA) and Ozow's internal policies.
• **Documentation:** Creating detailed documentation for each agent, outlining its functionality, limitations, and any necessary human oversight points.
• **Process Adherence:** Following Ozow's established change management, release, and security protocols for all AI feature deployments.
• **User Adoption:** Supporting business users in adopting new agent tools through clear guides, demonstrations, and ongoing assistance.
• **Continuous Learning:** Staying abreast of the latest developments in agent frameworks, Model Context Protocol (MCP), and new model releases, and sharing relevant knowledge with the team.
• **Process Identification:** Identifying manual business processes that are prime candidates for automation.

What They're Looking For

To excel in this role, you will need a strong foundation in software engineering and a keen interest in artificial intelligence. Essential qualifications and experience include:

• A degree in Computer Science, Engineering, or a closely related field, or equivalent practical experience.
• Proficiency in Python, coupled with comfort in working with REST APIs and JSON data structures.
• Hands-on experience building with at least one Large Language Model (LLM) API, demonstrated through personal projects, a GitHub repository, or prior professional work.
• A foundational understanding of RAG, embeddings, and prompt engineering principles.
• Proficiency with Git for version control and working knowledge of Docker for containerisation.
• A diligent and security-conscious approach to data handling, understanding the critical importance of protecting personal and financial information.
• Exposure to an agent framework such as LangGraph, CrewAI, or an agent SDK, or familiarity with MCP.

Additional desirable skills and experience include:

• Experience with Amazon Web Services (AWS).
• Familiarity with C# or .NET, given that a significant portion of Ozow's platform is built on ASP.NET Core.
• Working knowledge of SQL, specifically SQL Server or PostgreSQL.
• A demonstrated interest in the payments, fraud, or broader financial services sectors.

This role offers an exciting opportunity to contribute to cutting-edge AI development within a leading South African fintech company. Roles like this often involve building sophisticated systems that directly impact business efficiency. While salaries vary significantly based on experience and specific responsibilities, software engineering positions in the fintech sector in South Africa can offer competitive remuneration. Common employers for such roles include established financial institutions, innovative startups, and technology service providers. The career path can lead to senior engineering roles, AI specialisation, or management positions within tech departments.

• A degree in Computer Science, Engineering, or a related field, or equivalent practical experience.
• Solid Python, and comfort working with REST APIs and JSON.
• Hands-on experience building with at least one LLM API, shown through a project, GitHub repo, or production work.
• Basic understanding of RAG, embeddings, and prompt engineering.
• Git, and working knowledge of Docker.
• A careful approach to data: you understand why personal and financial data needs protecting.
• Exposure to an agent framework (for example, LangGraph, CrewAI, or an agent SDK) or to MCP. (Nice to have)
• AWS experience. (Nice to have)
• C# or .NET, since much of our platform runs on ASP.NET Core. (Nice to have)
• SQL (SQL Server or PostgreSQL). (Nice to have)
• Interest in payments, fraud, or financial services. (Nice to have)