Join the team redefining how the world experiences design at Canva.
The role is remote and available across Australia and New Zealand (ANZ).
You will be building enrichment solutions to tag and classify various content types in Canva's library.
The position involves designing data pipelines to automate metadata backfills for the existing content library.
You will be responsible for wrangling datasets to train and evaluate machine learning models.
Troubleshooting and resolving technical issues within the team’s domain will be part of your duties.
You will design and build bespoke machine learning solutions for novel problems such as style transfer and agentic design generation.
Proposing methods to decrease the cost of inference in the enrichment pipeline is expected.
You will manage stakeholders and identify cross-team collaboration opportunities to drive new ideas and improvements.
Investigating and providing guidance on interacting with the service and codebase in response to stakeholder requests is also part of the role.
The team consists of experienced engineers, including three senior machine learning engineers, three senior backend engineers, a senior backend architect, a product manager, and an engineering manager.
Requirements:
You must have more than 5 years of hands-on experience in designing and developing complex machine learning models, especially in computer vision.
Experience in R&D and conducting literature reviews on the latest machine learning techniques is required.
Proficiency in PyTorch and setting up cloud machine learning infrastructure is essential, with familiarity in large language models (LLMs) and prompt engineering as a must.
Familiarity with embeddings and vector databases is necessary.
You should have experience working with microservices and large monorepos.
Following disciplined coding practices, actively participating in code reviews, and setting best-practice standards for peers is expected.
Strong written and verbal communication skills and excellence in team collaboration are required.
You should take the time to fully understand problems before diving into code.
Preferred qualifications include holding a Master’s or PhD degree in a machine learning discipline.
Experience with Ray, Weights & Biases, Kubernetes, and Java is preferred.
Experience hosting LLM architectures and fine-tuning them through reinforcement learning is also preferred.
Benefits:
Equity packages are offered to ensure that your success aligns with the company's success.
An inclusive parental leave policy supports all parents and carers.
An annual Vibe & Thrive allowance is provided to support your wellbeing, social connection, office setup, and more.
Flexible leave options empower you to take time to recharge and support your personal needs.