Machine Learning Engineer

About the job

icetana is an AI solution that revolutionises the way cities, shopping centres and venues keep people safe. We’re a startup but we have some big customers like the MAF shopping centres in the Middle East and a number of Australia’s largest universities.

The role is for a machine learning engineer who is passionate about applying AI to vision use cases. This is an opportunity to collaborate and own the development of ML solutions that create a magical experience for our users. We have large ambitions with our product roadmap, so many of these projects will be taken from 1 to N. The person who fills this role will work closely with the rest of the development team; around 7 people.

We have a flexible working environment with a mix of in-office collaboration but with the flexibility to work from home if it makes sense. We’re in a cool office in WeWork and are listed on the ASX which adds a bit of excitement some weeks.

Everyone in the team is quite autonomous and scopes out their own work and requirements, the leadership team sets the goals and problems to solve for the cycle. We have a very helpful, collaborative and supportive environment where we talk a lot.

The machine learning pipelines are written in Python and we use PyTorch and OpenCV for model/algorithm development, use Docker and PDM for dependency management and productionise with Triton and TensorRT. We use PostgreSQL as a feature store, as well as several other tools for data storage, labelling and maintenance.

We’re a loud bunch who love a laugh. We believe in sustainable work since we see it as a marathon not a sprint. We go to the common WeWork area on Wednesday from 3-5 to have a chat amongst other startups while we finish off work. You’d be working with people who are the best in the field and people who have experience building large, successful businesses before. We value excellent technical skills, but also people who communicate and collaborate well.

Responsibilities

• Developing and improving a range of vision models including OCR, object detection and classification models.
• Creating algorithmic solutions to improve pipeline accuracy.
• Dreaming big, experimenting with and productionising VLMs.
• Productionising machine learning pipelines.

Qualifications

We honestly don’t care about qualifications, as long as you can do good work and are eager to learn.