AI / LLM product engineer
Summary
Focuses on applied AI by building and testing small language models on resource-constrained hardware like Raspberry Pi. The role involves Python-based prototyping, model quantization, and performance evaluation to determine the practicality of AI product ideas.
Salary: $150,000 – $180,000 per year
AI Product Engineer — Small Models & Applied AI
Melbourne - Mon, Wed & Thu in office
SuperCoders is recruiting an AI Product Engineer for a hands-on applied AI role.
The work is focused on finding out what AI models can do in practice when compute and memory are limited. You will take ideas, build prototypes, run tests, analyse the results and work out whether an approach is good enough to take further.
You’ll be working across software engineering and data science, including:
Building and testing small language models
Running real datasets through models and analysing the results
Writing Python tools for experiments, testing and evaluation
Comparing models, prompts and different approaches
Quantising and optimising models for lower memory and compute use
Measuring accuracy, latency and resource usage
Building repeatable ways to evaluate model output
Using larger models to help assess the performance of smaller ones
Moving experiments from development machines onto smaller, more constrained devices
Producing working prototypes and clear technical findings for engineering and product teams
You may be given a product idea or technical problem and asked to determine whether it is practical, what the limitations are and what would need to change to make it work.
We’re looking for someone with strong Python skills and solid technical depth in backend engineering, machine learning or data science. Experience with Linux is valued. So is experience working with small computing devices such as Raspberry Pi, embedded Linux systems or other resource-constrained hardware. You should be comfortable working directly with models, data, operating systems and hardware rather than only integrating AI APIs into applications.
Experience with small or local models, model evaluation, quantisation or optimisation would also be useful.
Tell us what you have built
We’re particularly interested in what you have actually finished and made work. AI and LLMs have made it much easier to turn ideas into working software. We therefore expect strong candidates to have a substantial list of things they have built, tested, shipped or completed - whether professionally or in their own time. Personal projects or commercial are both great - tell us about what you have done. We care about evidence that you build things, finish them and learn from the result.
Problem solving, critical thinking, curiosity and drive are important for this role. The selection process will include practical problem-solving exercises.
APPLY NOW! Via this advertisement or to apply@supercoders.com.au