The PEGroup is scaling automation and AI across payroll, public sector and SaaS. We need a Lead AI Automation Engineer to own the automation roadmap end-to-end — not simply to build what they are handed, but to decide what gets built, ship it, and prove the return. This person designs and delivers AI-assisted automations, internal tools, websites and backend services, and makes our business systems — CRM (HubSpot/Zoho), finance, collaboration and internal payroll/contractor platforms — operate as one connected system.
The gap this closes: automation ideas currently queue behind other priorities and depend on someone else to scope, sequence and champion them. This hire takes the problem, defines the solution, builds it, measures it, and raises the standard for everyone else doing automation work.
Must-Have
- Ownership and initiative. A track record of spotting the problem, scoping it, and shipping it without being managed into it. Comfortable telling the business what should be built and why.
- Automation delivery at depth. Production-grade workflows in n8n (self-hosted or cloud) or an equivalent orchestrator, including AI agent steps, retries, idempotency, queuing and human-in-the-loop approvals.
- Ships real software, not just flows. Backend APIs and services in Node.js/TypeScript or Python, and working websites/front-ends deployed and maintained in production.
- Fluent with AI development tools — Claude / Claude Code, Codex, Cursor, Lovable or equivalents — with shipped work to show for it. We treat AI-assisted delivery as the default way of working, not a novelty.
- Practical LLM engineering. Prompt design, structured outputs, tool/function calling, retrieval (RAG), evaluation of model output, and the judgement to know when a deterministic script beats a model.
- API integration depth. REST, GraphQL, OAuth2, webhooks; pagination, rate limits, idempotency keys, retries and graceful failure. Confident reading unfamiliar API docs and getting to a working call fast.
- Custom software and CRM systems. Configuring, extending and integrating platforms such as HubSpot and Zoho (CRM/Deluge) — objects, pipelines, custom fields, native workflows and their APIs — and knowing when to extend a platform versus build outside it.
- Data handling. JSON and SQL, data mapping and transformation, and system-of-record discipline (clarity on which system owns which field).
- Security and compliance mindset. Safe handling of PII, secrets and credential management, least-privilege access, and auditable execution.
- Engineering fundamentals. Git, code review, CI/CD, and environment promotion between dev and production.
- Communicates with non-technical stakeholders. Can run a discovery conversation with Finance, Ops or a client-facing team and return with a workable spec — and can explain trade-offs without jargon.
Nice to Have
- Experience leading or mentoring engineers, or being the first technical hire in a function.
- Docker and Linux for self-hosted infrastructure; practical DevOps (monitoring, alerting, backups, cost control).
- Vector databases and RAG at scale, agent frameworks, or MCP-based tool integrations.
- Finance and operations stack exposure: Xero, Wise, Airtable, Notion, Slack, Google Workspace, Microsoft 365.
- Domain exposure to payroll, umbrella, recruitment tech or contractor onboarding.
- Rapid prototyping and design sense — able to put a usable interface in front of a stakeholder in days, not weeks.
Location: Remote (global), with meaningful overlap to WAT/UK working hours.
Contract: 6 months (project temp), with strong potential to convert to permanent.
Team: Leads the automation practice; works alongside an Automation Specialist and a Zoho Developer; reports to the Head of Technology (Engineering).
AI tools (licensed and expected to be used daily): Claude / Claude Code, Codex, Cursor, Lovable, and comparable AI development tools.
Automation & orchestration: n8n (self-hosted and cloud), with Make/Zapier where appropriate; webhooks, scheduled jobs and queues.
Build stack: Node.js/TypeScript, Python, REST/GraphQL APIs, modern web front-ends, Docker/Linux, Git and CI/CD.
Business systems: HubSpot, Zoho (CRM/Deluge, Projects), Xero, Wise, Airtable, Notion, Slack, Google Workspace, Microsoft 365, LinkedIn/social APIs, and internal payroll/contractor systems.
Data & AI: LLM APIs, structured outputs and tool calling, vector search/RAG, evaluation frameworks, and MCP-based integrations.
Security: NDA required; least-privilege access, secret management and audit logs. AI tool use must follow our data-handling policy — no client PII into unapproved models.
Applying: we would rather see what you have built than read about it. Include links to live automations, repositories, sites or products, and be ready to walk us through one in detail.
Pay: R100 000,00 - R200 000,00 per year
Work Location: Remote
