AI Product Engineer

About the company

Mathspace is dedicated to improving mathematics education through technology. The team is building an adaptive, intelligent learning platform that helps students receive the right support at the right time so they can make real progress in math.

For more than ten years, the company has created interactive, feedback-based learning experiences that are already used by large numbers of students and teachers around the world. The organization is now expanding its AI capabilities to create personalized, real-time guidance that feels more like working with a tutor.

The culture is mission-led and combines educational understanding with modern software engineering. Work in this role has a direct effect on how future learners experience mathematics.

Role overview

As an AI Product Engineer, you will drive the creation of AI-enabled learning experiences built on large language models and other generative AI approaches. You will take ideas from early concept through production release, developing features that give students timely hints, conversational responses, and scalable math support.

You will be part of a cross-functional product team that includes engineers, designers, educators, and product managers. The role involves working across the full stack to explore how AI can be applied effectively at the intersection of pedagogy and product design. You will assess models, shape prompts and evaluation methods, and help deliver refined classroom-ready experiences.

This is an early-stage area of the business, so you will also help shape the company’s broader strategy, technical systems, and ways of working around LLM-based development.

Key responsibilities

• Build AI-driven learning features that bring LLM capabilities into student and teacher applications.
• Create, test, and ship LLM-based functionality such as subject-specific feedback, adaptive hints, automated question generation, and teacher assistant tools.
• Develop across the stack using React, TypeScript, Python, and GraphQL.
• Experiment with prompting, retrieval-augmented generation, vector search, and model fine-tuning to improve learning impact.
• Work closely with curriculum and pedagogy specialists to ensure the AI experience remains educationally effective and aligned with learning principles.
• Help improve AI infrastructure and LLM operations by contributing to evaluation pipelines, prompt version control, and monitoring/observability tools.
• Apply careful judgment to what AI should and should not do in an educational setting, and build appropriate safeguards.

Requirements

• At least 5 years of professional experience in product-oriented full-stack or backend engineering.
• Strong working knowledge of React, TypeScript, Python, and GraphQL.
• Practical experience building and running LLM-based products in production environments.
• Familiarity with retrieval-augmented generation, prompt engineering, evaluation methods, and LLMOps practices.
• Good product thinking and empathy for users, particularly in educational use cases.
• Strong communication and teamwork skills for cross-functional collaboration with pedagogy, design, and engineering.
• Motivation to work quickly, keep learning, and build work that meaningfully helps students and teachers.

What the company offers

• Purpose-led work focused on improving mathematics education for learners and educators.
• Opportunity to create impact at scale, reaching hundreds of thousands of learners across Australia, the US, and other regions.
• A small, trusted team structure where individual ownership and contribution matter.
• Remote-friendly working arrangements with some overlap with Sydney working hours; the head office is located near Central Station.
• Annual budget for learning, training, and professional development.
• 2.5 paid volunteer days each year through a partnership with pledge1percent.org.
• Competitive compensation plus equipment support, including a MacBook Pro and any additional peripherals needed for effective work.

Additional information

The role is based in Australia and supports remote work. Candidates should be able to overlap with Sydney time zone for collaboration.