Head of Agentic Operations (AI Workforce Architect)

Remote Choice; South Africa; Remote; ZAR 1,800,000 - 3,000,000; Full time

Job summary

Remote Choice is assisting a UK-headquartered group to appoint a Head of Agentic Operations (AI Workforce Architect) remotely for South Africa. You will design, build, govern, and optimise a digital workforce of AI agents and automated workflows across functions including sales, broking, compliance, finance and HR.

As a hands-on leader, you will drive cost savings, hours removed, and measurable business value while ensuring governance and data protection across jurisdictions.

Qualifications

• At least 5 years building software or automation; includes 2 years building production-ready LLM-based agents/workflows.
• Deep experience with at least one automation platform (Make, n8n, Zapier, or Power Automate).
• Practical experience in prompt design, tool usage (function calling), and evaluating model outputs.
• A portfolio of deployed agents or automations with measurable results.
• Clear written English and ability to explain technical work to non-technical colleagues.

Responsibilities

• Design and build AI agents and automations using Claude and other LLMs via APIs and connectors.
• Develop and maintain prompts, tool definitions, connectors, and scheduled tasks.
• Prototype solutions rapidly and prepare production-ready implementations.
• Maintain the Group Agent Register with owners, data access, schedules and performance measures.
• Conduct weekly audits and retire agents that fail or underperform.
• Ensure human approval gates and least-privilege access for security.
• Report monthly on hours removed, costs avoided, and income supported by agents.
• Train staff to work with AI agents and provide operating procedures.
• Control automation tool costs and document the environment.
• Target KPIs: complete agent register, deploy production agents, demonstrate value.

Skills

Development Experience

Automation Platforms

LLM Development

Proven Delivery

Communication

Tools

Make

n8n

Zapier

Power Automate

Claude API

Microsoft 365

Job description

Head of Agentic Operations (AI Workforce Architect)

Remote, South Africa | Posted on 27/10/2026

Head of Agentic Operations (AI Workforce Architect)

Senior Leadership Opportunity | Insurance, Financial Services & Investment Group

Remote Choice is recruiting on behalf of its client , a UK-headquartered international investment group operating across the insurance, financial services and property sectors, for an experienced Head of Agentic Operations (AI Workforce Architect) .

This is a senior leadership opportunity for a hands‑on AI and automation specialist to drive operational transformation within an international, technology-focused business.

Position Overview

Job Title

Head of Agentic Operations (AI Workforce Architect)

Level

Location

Remote, South Africa

Working Hours

UK Business Hours

Reporting To

Key Working Relationships

Functional Leads, Group Salesforce Engineer, IT Support and Executive Owner for AI Governance and Risk

Employment Arrangement

Employed through Remote Choice as Employer of Record (EOR) on behalf of the client Group

Purpose of the Role

The Head of Agentic Operations will design, build, operate and govern the Group's digital workforce, comprising AI agents and automated workflows across sales, broking operations, compliance, finance and HR.

The objective is to automate routine tasks, enabling employees to supervise outcomes and focus on clients and higher‑value activities.

This is a hands‑on role requiring a builder‑first, manager‑second approach, with success measured by hours removed, costs avoided, and income supported by AI agents.

Key Responsibilities

AI Agent Development

Design and build AI agents and automations using Claude and other large language models (LLMs), Make, monday.com, Zendesk, Salesforce, Xero, and Microsoft 365 through APIs and connectors.

Develop and maintain prompts, tool definitions, connectors, and scheduled tasks to a consistent specification.

Prototype solutions rapidly and prepare successful implementations for reliable production use.

Agent Management and Monitoring

Maintain the Group Agent Register, documenting each agent's owner, purpose, data access, schedule, controls and performance measures.

Conduct weekly audits, improving, fixing, or retiring agents that fail or do not deliver sufficient value.

Monitor agent runs, errors, and operating costs, taking corrective action where necessary.

Governance and Data Protection

Implement human approval gates for all consequential actions performed by AI agents.

Apply least‑privilege access, logging, testing and evaluation before production release.

Ensure compliance with UK GDPR and South Africa's Protection of Personal Information Act (POPIA), working with the Head of Compliance and the executive responsible for AI governance.

Performance and Business Value

Report monthly on hours removed, costs avoided, error rates, and income supported by each agent and business function.

Prioritise development based on measurable commercial and operational value.

Staff Training and Adoption

Train employees to work effectively with AI agents, including through the Group's monday.com and Claude competency programme.

Develop clear operating procedures, ensuring every agent has a named human owner who understands its operation.

Technology and Cost Management

Select appropriate automation tools and vendors while controlling subscription and usage costs.

Maintain documentation of the automation environment to ensure operational continuity without reliance on one individual.

Key Performance Measures

Performance will be assessed from the first quarter, with specific targets confirmed in writing during the first week.

Performance Area

Target

Agent Register

Complete the Group Agent Register with named owners and weekly audits in place by day 30.

Production Agents

Deploy at least 5 new or rebuilt agents into production, each with a named owner and defined measures, by day 90.

Business Value

Report measured hours removed and costs avoided monthly from month 2, against an agreed target.

Reliability

Maintain an agent run success rate above 95%, with failures resolved within 2 working days.

Safety

Zero data protection incidents caused by agents, with human approval for every consequential action.

Staff Adoption

Achieve the agreed staff certification target through the competency programme by month 6.

Essential Qualifications, Skills and Experience

Development Experience: At least 5 years building software or automation, including 2 years developing LLM-based agents or workflows used in production.

Automation Platforms: Deep experience with at least one platform, such as Make, n8n, Zapier, or Power Automate.

LLM Development: Practical experience in prompt design, tool use (function calling), and evaluating model outputs.

Proven Delivery: A portfolio of deployed agents or automations with measurable results.

Communication: Clear written English and the ability to explain technical work to non-technical colleagues.

Desirable Qualifications, Skills and Experience

Experience with the Claude API, Model Context Protocol (MCP), Claude scheduled tasks or skills.

Integration experience with monday.com, Salesforce, Zendesk, Xero or Microsoft 365 (Graph).

Insurance or financial services experience, including an understanding of regulated data.

Experience leading a small technical team or contractors.