Summary
A mid-level AI engineer will join an early-stage team to build production AI systems for businesses across multiple sectors. The role combines hands-on software engineering with AI-focused development, offering substantial influence over tooling, architecture, development practices, and the full journey from planning to production.
Highlights
Mid-level role with unusually high ownership and direct collaboration with the founder. You can shape tooling, architecture, and development strategy while building production AI systems for real-world business problems across multiple sectors.
Description
About the role
Sentry AI is hiring an Artificial Intelligence Engineer to help build production AI for businesses across ANZ and Asia.
This is a mid-level role, and we care far more about what you have built and how fast you learn than the years on your CV.
This is our first hire, so you will work directly alongside the founder and see how the whole business runs.
You get real control over the development strategy: the tooling, the architecture and the way we build are yours to shape.
We want an ambitious developer with real experience on the front end, the back end, or both, and a genuine obsession with AI.
The role has two halves: shipping AI systems for our clients, and using AI to run your own work, from planning a build through to watching it live in production.
We build for businesses that are counting on it, across healthcare, recruitment, real estate and SaaS.
The problems are state of the art, and they have real consequences when they are wrong.
What you will build
Take a rough client brief and turn it into a system that works, then own how it behaves once real people are using itBuild agents that do real work inside a business: retrieving, deciding and acting on live data, with a human approving anything that carries a consequenceShip voice AI that holds a real conversation with a real customer, including the awkward paths where it does not go to planDirect AI agents through most of the build: set the architecture and the context, review what comes back, and catch the wrong answer before it shipsRun our in-house WATCH methodology on live projects, and sharpen it as you go
The stack
What we run today.
You will not know all of it on day one and nobody expects you to.
Frontend.
Next.js, TypeScript, Vercel.
The interface over your own work.Backend.
Python, FastAPI, Supabase and Postgres.
Where the services and the agents run.Data.
Supabase, AWS RDS, RAG, vectorisation and embeddings.
How a model reaches a company's own knowledge.AI.
Claude, OpenAI, Gemini, MCP, LangChain, Langfuse, AgentCore, OpenClaw, frontier and open-weight models.
The agentic layer: how a model gets context, tools and limits.
How we work
Agentic by default.
Coding agents, harnesses and context files are standard issue here.
The bar is not whether you use AI, it is how well you direct it and how quickly you notice when it is wrong.
The WATCH methodology.
WATCH is our in-house development methodology for AI-first engineering: how a rough client brief becomes an agentic system that is safe to point at a live business.
Five stages, and you will run all of them:
Workshop.
Plan it.
Read the client profile, organise the problem, and log the work to actions.mdArchitect.
Write task.md and test.md before a line of it gets builtTask.
Build to task.md, and write the tests as you goCheck.
Run the tests, react-doctor, and the security and compliance auditHand off.
Open the in-depth PR, watch CI, and queue the merge
What we look for
Passion projects over a decorated CV.
Show us the coolest thing you have built and what broke along the wayReal experience on the front end, the back end, or both.
Depth on one side and curiosity about the other is enoughAn AI-first mindset.
You build with coding agents daily, and we will ask you to walk us through your setupYou keep up: you already know what changed in the tooling this quarterThe right to work in New Zealand, and able to be in the Eden Terrace office two to three days a week
Nice to have
Experience with some of the tools in our stack aboveVoice AI experience: Retell, ElevenLabs, or telephony in generalSide projects we can look atNone of these are requirements.
If you match the list above and none of this one, still apply.
What you get back
NZ$84,000 a yearHybrid by design: 2 to 3 days in the Eden Terrace office, the rest from homeCurrent tooling, paid for.
We back your AI use rather than rationing itWork that ships: real systems in front of real users, not prototypesReal ownership: you shape the development strategy alongside the founder, and what you ship is yours to run
