About the role

Maincode is training Matilda, a large language model built and trained from scratch in Australia. Our new compute cluster is now live, and we are scaling the next version and deploying it publicly.

This role sits inside the production system that serves Matilda. You will build and maintain the back end services that make the model usable in the real world: APIs, infrastructure, and the systems that turn a trained model into a reliable public capability.

We build AI systems end to end. We design the architectures, run the infrastructure, train the models, and operate the systems ourselves. Matilda is not a research prototype. It is a production system trained at scale and served publicly.

Maincode operates one of the largest private AI compute environments in Australia, built for training and operating our own models. You will be working directly on the systems that deploy and serve a model trained from scratch.

What you would actually do

You will build and maintain the services that sit between the model and the outside world. This includes work such as:

• Building and maintaining services that handle model inference and user requests
• Designing systems that manage requests, sessions, and streaming responses
• Implementing reliability mechanisms such as rate limiting, retries, and graceful failure
• Building authentication and access controls for public usage
• Designing systems for logging, telemetry, and evaluation signals
• Improving latency, throughput, and reliability of model serving
• Integrating new model checkpoints into the production system
• Working closely with training and infrastructure engineers to deploy and operate the model

Much of the work happens inside production systems: logs, traces, performance profiles, and deployment pipelines. The goal is not polish. The goal is a system that stays up, stays fast, and behaves predictably under load.

The kind of person who does well here

We are looking for engineers early in their careers who want to learn how production AI systems are actually built and operated. You may have one or two years of experience building production software. What matters most is curiosity, reliability, and the willingness to learn how large scale systems behave under real constraints.

People who tend to do well here:

• Care about runtime behaviour and system reliability
• Enjoy debugging real systems rather than writing theoretical code
• Think clearly about system boundaries and failure modes
• Stay calm and methodical when production behaves unexpectedly
• Want to understand how large scale AI systems actually work

You do not need prior experience serving large language models. You do need the discipline to build systems that are hard to break.

How you would work

You will use code as a way of shaping a production system. You should be comfortable:

• Building back end services in a modern language (Python is common here)
• Working with APIs and service interfaces
• Designing systems that remain stable under load
• Reading logs and system metrics to understand behaviour
• Collaborating closely with training, infrastructure, and product engineers

Speed matters, but so does rigour. Reliability is a feature.

What this role is not

• It is not maintaining internal business software or conventional product back ends
• It is not integrating third party AI services or building on top of external models
• It is not primarily front end work or prompt engineering
• It is not incremental feature work on mature systems

This role focuses on building and operating the systems that deploy and run a model we train ourselves, where the core problems are performance, scale, and reliability.

Why Maincode

Maincode builds AI systems end to end. We train the models, run the infrastructure, and operate the systems ourselves. You will work with a small team that:

• Builds the full AI stack rather than outsourcing it
• Treats reliability and system design as core engineering problems
• Values engineers who want to understand how systems actually work
• Is building long term capability in training, deployment, and serving

If you want to work directly on the systems that deploy and operate a large language model trained from scratch, this role will put you inside that work.

Note

This is a full time role based in Melbourne, working closely with our in person engineering and research team. At this time we are not able to offer visa sponsorship, so applicants must have existing and unrestricted work rights in Australia.