Engineer

ELab Ltd.; New Zealand; Hybrid; NZD 120,000 - 180,000; Full time

Job summary

ELab Ltd. in New Zealand or Australia offers a role to design and deliver production software for clients. You will own technical decisions, work with the CTO, and drive deployment, handover and ongoing improvement.

You’ll build AI-enabled applications, choose architectures, and ensure secure, scalable solutions while collaborating with advisory teams on discovery and client requirements.

Qualifications

• Built and operated production applications, including post-release support.
• Strong Python skills and depth in another production language.
• Experience with LLM APIs, tool calling or agent workflows.
• Ability to deploy cloud infrastructure with identity, networking, and security considerations.

Responsibilities

• Build production-grade applications, AI agents and integrations for clients.
• Make architectural choices and explain trade-offs to stakeholders.
• Design access controls and data boundaries for sensitive environments.
• Turn delivery lessons into reusable components and practices.

Skills

Python

Cloud infra

Data engineering

LLM integration

Tools

LLM APIs

Job description

Bring your engineering experience to AI delivery. Design and build production software with clients, own the technical decisions and keep improving it.

New Zealand or Australia. Remote or hybrid.

You’ll work alongside our CTO and co-founder, Franco Sioquim, to turn client requirements into working software. Our advisory team leads discovery. You take the technical work forward, designing the solution with the client and owning it through deployment, handover and ongoing improvement.

What You’ll Do

Build for Production

Develop applications, AI agents and integrations that work with a client's existing tools. Take responsibility for testing, deployment, monitoring and the production issues that follow.

Make the Technical Calls

Choose the architecture, model and cloud environment to suit the work. Explain trade-offs in reliability, latency and cost to both technical and business stakeholders.

Protect Client Information

Design access controls and data boundaries for sensitive environments. Work through identity, permissions and regional hosting requirements before deploying.

Understand the workflow your software supports. Present your approach, test it with the people using it and leave them with clear documentation and a working handover.

Improve How We Build

Turn lessons from delivery into reusable components and engineering practices. Use AI development tools deliberately, review the code and verify the result.

What You’ll Bring

• You’ve built and operated production applications, including maintaining them after the initial release.
• You bring strong Python skills and depth in another production language, with experience across application development, data engineering or cloud infrastructure.
• You’ve built with LLM APIs, tool calling or agent workflows, and understand how to evaluate output quality and handle failure.
• You can deploy cloud infrastructure and reason about identity, networking, security and cost. We aren’t looking for one particular cloud vendor.
• You use AI in your engineering workflow and can explain how you review, test and take responsibility for its output.
• You’re comfortable explaining technical choices to clients and taking a piece of work through to completion.

Useful Experience

These would give you a head start. You don’t need every one.

• MCP servers, retrieval over business documents and data pipelines.
• Microsoft 365, SharePoint or Power Platform integrations.
• Financial services, healthcare or other environments with sensitive information.
• Consulting, forward deployed engineering or products you’ve built independently.

The First Conversation

We’ll start with a conversation about your experience and what you’re looking for. The engineering process includes a practical technical exercise, scoped to around two hours, followed by a discussion of your approach and a conversation with the team.