Senior Software Developer

Senior Software Developer - Product Engineering & Agentic AI
Penrose, Auckland | Permanent, full-time

We're building a new platform within Freightways' Express Package group. It will run like a startup: a small team, short feedback loops, and decisions made close to the work, backed by an NZX-listed group.

The platform will replace several of the group's bespoke applications. We're building it from the ground up, then moving customers across in stages as it becomes ready for each migration. We'll start from what customers and operational teams actually need rather than reproducing what the old systems did, and retire each application once its customers have moved.

AI agents are central to how we'll do this. They make it practical for a small team to build a new platform, understand what the existing applications do, and migrate customers and data safely. Most of our development doesn't work this way yet; this project is where we prove it. Engineers write the specification and acceptance criteria, direct agents through planning, implementation, testing and documentation, and review everything before it ships. The engineer owns the result, regardless of what wrote the code.

We're hiring a senior developer who already works this way and can help set the standard the rest of our teams will adopt.

The work

Freightways Information Services is the IT shared-services business for Freightways Group. You'll build and support the platform used by customers and operational teams across NZ Couriers, Post Haste and the wider Express Package group. The stack is .NET and C#, TypeScript and React, and SQL, running on AWS and Azure.

You'll work with our product team to understand the customer problem before anyone writes a spec. You'll then take the work through build, release, migration and support in production. Part of the job is time in depots and at customer sites. You don't need a freight background, but you will need to learn the operation, because the software has to fit it.

How we'd like you to work

You'll pick up problems without waiting to be assigned them: work out what's needed, propose an approach, and see it through.

Because this project is setting the pattern for our wider development teams, a large part of the role is making others effective. You'll help shape the agentic workflow, write the shared agent instructions, skills and verification checks, and show other developers how it works in practice. You'll also unblock people and review their work properly.

When there's a disagreement, argue the case on its merits, then commit to whatever the team decides.

What you'll bring

• Production experience delivering software with AI coding agents, including the specifications, tests and verification controls that make it safe
• A commercial track record of taking systems into production and supporting them in service
• Full-stack skills across .NET/C#, TypeScript with a modern frontend framework (ideally React), and SQL data modelling
• Cloud delivery on AWS and/or Azure, with infrastructure as code, containers and CI/CD pipelines you've built and owned
• Secure application design: authentication, authorisation, secrets handling, and automated testing at every layer
• The ability to explain a technical design to non-technical people, and to learn a new business quickly

Useful but not essential:

• Experience building a new platform and migrating customers onto it from legacy systems in stages
• Depth in observability tooling
• Experience building agent tooling or evaluation harnesses
• Mentoring or coaching experience
• A background in logistics, e-commerce or another operationally demanding industry

Conditions

This is an office-based role in Penrose. You'll be on a shared on-call roster with the other engineers. Some releases, cutovers and customer migrations are scheduled outside business hours to avoid disrupting operations.

Applying

With your application, tell us about something you've delivered using AI agents: what you asked the agents to do, how you checked the output, and what you'd do differently. A link to a repo, pull request or spec is welcome, and you'll walk us through it at interview. You're welcome to use AI to help with your application, as long as you can explain and defend everything in it.