3–6 Month Contract | Auckland, New Zealand | MacroActive
Build the AI systems that will transform how a business operates.
We're not looking for someone to experiment with AI. We're looking for someone who knows how to make it work in the real world.
At MacroActive, we're building an AI Agent Factory - a system of intelligent agents that can write and test code, analyse business data, build marketing campaigns and automate complex work that people currently do manually.
We need an experienced AI builder who can take this vision and turn it into a working, measurable reality.
This is a hands-on, high-impact contract for someone who enjoys solving difficult problems, making independent decisions and building systems that deliver real results.
If you've already built AI agents that people rely on, and you're ready to build something bigger, we'd love to hear from you.
About MacroActive
MacroActive is a full-service agency and technology platform helping fitness and wellness creators grow their businesses online. We provide the technology, content services and lead generation support that help creators turn their expertise into thriving businesses.
Our platform supports creators across more than 30 countries, and our creators have earned over $220 million through it.
Our business model is simple: we succeed when our clients succeed.
Now, we're taking the next step in how we operate - using AI agents to make our business smarter, faster and more efficient.
The opportunity
Imagine a system where a request for a new feature can be turned into code, tested, reviewed and prepared for release by AI agents, with humans providing direction and making the important decisions.
Imagine agents that can analyse business data, identify opportunities, build campaigns and continuously improve their own results.
That's what we're building.
As our Agent Factory Lead, you'll design, build and improve the systems that make this possible. You'll establish the architecture, decide how agents work together, measure their performance and ensure they are reliable, secure and cost-effective.
You won't simply be using AI tools to get your own work done. You'll be building the systems that enable an entire business to get more done with AI.
This is a contract with genuine ownership, where your work can have a direct and visible impact on the business.
What makes this opportunity different?
Real problems. Real data. Real outcomes.
You'll work with valuable business data, including hundreds of thousands of coach and client conversations, tens of thousands of exercise videos and more than a decade of data connecting social content to sign-ups and revenue.
You build the system, not just the features.
Your focus will be on creating the agents, tools and processes that produce work reliably - rather than manually building every feature or completing every task yourself.
You'll have room to make decisions.
There is no established playbook to follow. You'll help define the architecture, standards and processes, test your ideas and demonstrate what works.
Results matter more than impressive demos.
We care about whether agents produce accurate results, how much they cost, how quickly they work and whether people can trust them. You'll have the opportunity to measure all of this and improve it continuously.
What you'll be responsible for
1. Build our AI agents
Develop agents that can:
Turn defined software requests into working code, with the right context and safeguards.
Write and run tests, review code and identify problems before changes are released.
Analyse business data and produce accurate, useful insights.
Support marketing campaigns and automate operational tasks.
Work together to complete larger tasks that would otherwise require multiple people and steps.
2. Build the system behind the agents
Create the foundations that allow agents to work reliably at scale, including:
Tools and integrations: Connect agents to APIs, business systems and other tools, with appropriate permissions and safeguards.
Workflows: Design how agents work together, manage tasks and hand work over when human input is needed.
Knowledge and data: Help agents find and use the right information, and verify that their answers are accurate.
Testing and evaluation: Build ways to measure whether agents are doing the job correctly and catch problems when changes are made.
Monitoring and cost control: Track performance, reliability, speed and cost, and identify where improvements are needed.
Security and oversight: Ensure agents operate within clear boundaries, with human approval where necessary and sensible fallback processes when things go wrong.
3. Make the factory better every week
You'll measure what the agents produce, identify bottlenecks and improve their performance.
You'll look at questions such as:
How much work can the agents complete successfully?
How often do humans need to fix or redo their work?
Are mistakes reaching production?
What does each completed task cost?
How can we improve quality without unnecessarily increasing cost or complexity?
You'll also help the wider team adopt these new ways of working, explaining your decisions clearly and demonstrating the value through results.
Who we're looking for
We want an experienced, practical AI engineer who has already moved beyond prototypes and can demonstrate that they've built and launched working systems.
Essential experience
Real-world AI agents: You've built and deployed at least one agentic system that real users depend on. You can explain what it does, how you measured its success and what went wrong along the way.
AI-powered software development: You've worked with coding agents that can understand a codebase, write code, run tests and review changes. You understand where they work well and where human judgement is still essential.
Strong software engineering: You can write and maintain production-quality software using TypeScript and/or Python, with experience in APIs, databases, automated testing and deployment.
Working with AI models: You're comfortable with model APIs, tool calling, structured outputs and giving models the right instructions and information to complete tasks.
Testing AI properly: You know that an agent sounding convincing doesn't mean it is correct. You've developed ways to test its outputs, track failures and improve reliability.
Data and information retrieval: You have practical experience helping AI systems find and use relevant information, including retrieval-augmented generation (RAG), search or similar approaches.
Reliable systems: You understand how to manage failures, protect data, control costs and make systems dependable when real users and workloads are involved.
Independent problem-solving: You can take an unclear problem, work out what matters, make sensible decisions and deliver results without needing every step mapped out for you.
Driving change: You've introduced a new system or way of working and helped people adopt it, even when there was initial resistance.
Experience that would be a bonus
Building systems where multiple AI agents work together.
Using AI to analyse business data and generate insights that can be independently checked.
Improving AI prompts and workflows using measured results rather than guesswork.
Experience with knowledge graphs, advanced search or automated optimisation.
Leading, mentoring or setting technical standards for other engineers.
Experience building systems that handle high volumes of users or sudden spikes in demand.
You don't need to tick every bonus box. What matters most is that you've built real systems, understand their limitations and know how to make them better.
What this role is - and isn't
This role is about building the machinery that gets work done.
It isn't simply writing prompts, experimenting with AI tools or developing individual product features. It isn't about producing an impressive demo that falls apart when real users depend on it.
We need someone who cares about the entire journey: from understanding the problem to building the system, testing it, measuring its performance and improving it.
You'll need to be comfortable working with uncertainty, making decisions and explaining your reasoning. We value people who can stand behind their decisions, listen to evidence and change course when the facts show a better way.
What success looks like in your first two months
First three weeks: Understand how work currently gets done, identify the biggest opportunities and establish the initial testing, monitoring and cost-measurement processes for our coding agents.
By the end of six weeks: Have coding and quality agents operating within controlled boundaries, with clear measures for successful completion, rework and defects.
By the end of two months: Have a second area, such as business analysis or marketing, using the same foundations, with clear reporting on output, quality and cost.
These are the initial goals. We'll work together to agree on priorities based on the business's needs and what is achievable.
The contract
Location: Auckland, New Zealand.
Duration: Initial 3–6 month contract.
Focus: Hands-on AI engineering, agent development, systems architecture and automation.
Ownership: Significant responsibility for the design, delivery and ongoing improvement of the Agent Factory.
We're looking for someone who wants to make a meaningful contribution, not simply complete a list of tickets.
How to apply
We'd rather see what you've built and hear how you think than read a generic cover letter.
Please send your CV, GitHub profile or examples of your work, and brief answers to these four questions.
1. Tell us about an AI agent you've built that real people use. What does it do, how did you know it was working, and what was the biggest problem you had to fix?
2. How would you build a system of AI agents that takes a software request through to tested and reviewed code? Explain your approach in plain language and tell us what you would measure first.
3. An AI agent is producing good results, but its cost suddenly triples. How would you investigate what happened and decide what to fix?
4. Tell us about a time you introduced a new system or way of working that people were reluctant to adopt. What did you do, and what was the outcome?
Send your application and responses to [email protected].
If you're an AI builder who wants to move beyond individual agents and build the system that brings them together, this is your opportunity to make a real impact.
Come build the factory with us.