Agentic AI & Automation Engineer

Guardrisk is the undisputed market leader in cell captive insurance and risk solutions. We are renowned for our innovative approach to cell captive structures and other alternative risk transfer solutions for our clients. Guardrisk offers clients custom designed cover and is registered in South Africa as an insurer for all statutory classes of non-life and life insurance business.

Work directly with business users and Product Managers to understand and document business processes identify high-value automation opportunities and design prototype and engineer production-ready workflows automations and AI agents on Guardrisks governed platforms.

What makes this different

• Combines workflow automation software engineering and applied/agentic AI
• Works directly in the business context to map processes pain points decisions hand-offs and controls
• Identifies where to simplify automate or augment work rather than merely digitising an inefficient process
• Rapidly prototypes automations and agents with users and learns through deployment
• Designs agentic workflows orchestration human-in-the-loop controls and exception handling
• Connects business processes to systems APIs data models and enterprise services
• Builds reusable automation components agent skills tools and workflow patterns
• Makes sound design choices beyond one immediate use case with reuse across Guardrisk in mind
• Partners with Technology on platform integration security and productionisation
• Establishes the engineering practices documentation and patterns needed to grow Guardrisks AI & Automation capability
• Supports adoption and measures whether deployed automations actually improve the underlying process

Expected seniority: Energetic hands-on engineer with enough technical depth to engineer integrations workflows and agents but also enough business and product judgement to determine what should be automated where AI adds value where deterministic workflow is better and where human judgement or control must remain.

About the role

Were building a practical Agentic AI and Automation capability at Guardrisk and were looking for a hands-on engineer who wants to help shape it.

This is not a role focused on producing AI demos or implementing technology in isolation. You will work directly with business teams to understand how work gets done identify where processes can be simplified or automated and turn those opportunities into working solutions.

Youll combine business process thinking workflow automation software engineering and applied AI to build everything from straightforward process automations and integrations to AI agents and agentic workflows.

Just as importantly youll help us create the capability itself: reusable components patterns engineering practices and skills that allow successful solutions to be applied across Guardrisk.

Guardrisks Data & AI CoE is being designed around reusable capabilities and business-facing delivery with AI and automation capacity forming part of that model. COE Org Structure The wider approach also envisages agents working across enterprise systems while retaining human decision-making where appropriate. CoE Ways of Working Standard

What you will do

• Work directly with business users and Product Managers to understand map and document end-to-end business processes.
• Identify manual work bottlenecks repetitive activities and hand-offs that can be simplified automated or augmented using AI.
• Translate business problems into practical automation workflow and agentic AI solutions.
• Rapidly prototype with users test ideas in the real business context and turn successful prototypes into robust solutions.
• Design and build AI agents and agentic workflows that can reason over information use tools interact with enterprise systems and support business processes.
• Build traditional workflow automations where deterministic automation is a better fit than AI.
• Integrate solutions with enterprise systems APIs data and services.
• Design appropriate human-in-the-loop controls approvals exception handling and escalation paths.
• Develop reusable agent skills tools connectors workflow components and automation patterns rather than solving every problem from scratch.
• Work with Technology Architecture Security and other teams to take solutions from prototype through to controlled production use.
• Build evaluation and monitoring into AI and automation solutions so that we understand whether they are working as intended and delivering value.
• Document solutions patterns and lessons learned so that the capability becomes repeatable across Guardrisk.
• Help business and technical teams understand where agentic AI conventional automation integration or process redesign is the appropriate solution.
• Contribute to the standards and engineering practices that underpin Guardrisks emerging Agentic AI & Automation capability.

Minimum Qualifications

Were looking for someone who can operate across software automation and AI rather than specialising in only one of them.

• Bachelors degree in Computer Science Engineering Information Systems or a related technical discipline or equivalent practical experience.
• Strong software engineering experience in at least one modern programming language ideally Python C# JavaScript/TypeScript or similar.
• Practical experience building workflow automations system integrations or business process solutions.
• Experience designing and consuming REST APIs and integrating multiple systems or services.
• Hands-on experience building applications using large language models or generative AI APIs.
• Understanding of agentic concepts such as tool use orchestration retrieval structured outputs state and human-in-the-loop patterns.
• Experience taking solutions beyond prototypes into production or operational use.
• Ability to work directly with non-technical users to understand a business process and translate it into a technical solution.
• Strong engineering fundamentals including source control testing documentation security and maintainability.
• Ability to communicate technical trade-offs clearly to both business and technical stakeholders.

Preferred Qualifications

You dont need to have experience with every technology below. Were particularly interested in people who have gone deep enough in this space to understand what works in production and what does not.

• Hands-on experience developing AI agents multi-agent or agentic workflow solutions.
• Experience with agent frameworks orchestration approaches or modern AI engineering toolchains.
• Experience with enterprise automation technologies such as Microsoft Power Automate Power Platform Azure Logic Apps or equivalent workflow/orchestration platforms.
• Experience with the Microsoft Azure and Microsoft 365 ecosystem particularly where AI automation and enterprise data intersect.
• Experience applying techniques such as retrieval-augmented generation (RAG) tool/function calling structured extraction and AI-assisted workflow orchestration.
• Experience with process discovery process mapping BPMN Lean process improvement or similar approaches.
• Experience designing reusable automation platforms components libraries or engineering patterns rather than only delivering one-off solutions.
• Experience implementing evaluation observability or monitoring for AI applications.
• Understanding of responsible AI privacy security and appropriate controls for deploying AI into business processes.
• Experience in financial services insurance or another regulated environment.
• Evidence of being a builder and capability creator: someone comfortable starting with an ambiguous business problem working with users building something quickly and progressively turning it into a production-quality solution.

Core Outputs

• Documented business processes and automation opportunities.
• Production-ready workflow automations and integrations.
• AI agents and multi-step agentic workflows.
• Reusable agent skills tools connectors and automation components.
• Human-in-the-loop and exception-handling patterns.
• Working prototypes and MVPs.
• Process and solution designs.
• Evaluation monitoring and operational controls.
• Deployment-ready code and technical documentation.
• Reusable engineering patterns and standards for Guardrisk.
• Knowledge transfer and practical enablement that grows internal automation capability.