Software Engineer - AI Products

Your idea of a great day? Your eval suite catches a regression before a customer does. The prompt architecture you rebuilt last month is holding steady under real load. You've got real autonomy over how the system is built - and a team that trusts you to use it well. We've been looking for you. We're building Lantern, a new AI-powered fundraising copy platform, backed by over a decade of Donor Republic's fundraising expertise. It's a live commercial product entering the market with its first customers, not a prototype or an internal experiment. Every piece of copy Lantern generates needs a system behind it that keeps things accurate, on-brand and something a fundraiser can trust without double-checking. That system doesn't build itself. It needs someone who genuinely understands prompt architecture, evaluation and the guardrails that keep AI outputs consistent at scale - not someone bolting AI features onto a front-end build. We're looking for a Software Engineer – AI Products to take technical ownership of Lantern and the Assessment Hub, and to help shape what comes next as our AI product portfolio grows. This role is open nationally across Australia and New Zealand. Remote considered. About Donor Republic Donor Republic is Australia and New Zealand's leading full-service fundraising and marketing agency. We partner with charities and not-for-profits to help them raise more, connect better with supporters, and build lasting impact. Our AI, Education and Data team builds and operates Lantern and the Assessment Hub, alongside new commercial products as our AI portfolio expands. About the opportunity This is a live, revenue-generating product with real customers and real contractual obligations from day one. You'll own the backend and AI-systems engineering that makes it reliable: prompt architecture, evaluation frameworks, structured outputs and the guardrails that keep results trustworthy. You'll also maintain a clean front-end within our established design system - you won't be setting visual direction, but you will be building it well. You won't carry this alone. An engaged technical partner provides independent architecture and security audit, plus continuity cover, so the product has genuine resilience beyond one person. What you'll do Build and maintain prompt architecture, evaluation frameworks and output quality measurement for a commercial AI productDesign structured output schemas, guardrails and consistency logic that keep results reliableMaintain and extend our backend, database (Postgres/Neon) and API integrations (OpenAI, Anthropic, Stripe and others)Own our Google Cloud Platform environment (Cloud Run, IAM, Secret Manager, CI/CD, monitoring)Implement front-end features within our established design systemContribute to new commercial products as our AI portfolio expands What we're looking for We'd expect around 3-5 years of engineering experience, full-stack but weighted toward backend and AI systems, including at least 1-2 years of genuine prompt engineering or evaluation work on a live product with real users. This is a young specialism - we're not expecting five years of exactly this, but we are looking for real production depth, not demo or prototype experience. What you'll bring Genuine experience with prompt engineering, evaluation and LLM output quality in a production setting, not just prototypes or demosStrong backend fundamentals and API integration experienceSpecific Google Cloud Platform competence, and a clear sense of what should live on GCP versus a database or CDN providerCompetent front-end engineering (React/TypeScript) - you don't need to originate design, just implement it well, though you're welcome to bring ideas to the tableThe judgement to raise considered questions on a brief during discovery, and the reliability to deliver against it once scopedComfort working alongside an engaged technical partner and incorporating independent audit findings, without treating it as a challenge to your autonomy Why Donor Republic You'd be joining an established, respected agency backing a genuinely new product function. Real technical ownership, a commercial product entering the market with its first customers, without the risk profile of an early-stage startup. How to apply The full position description is available here: https://www.dropbox.com/scl/fi/gffl09c7zj5qn8w21v0eg/Software-Engineer-AI-Products-PD-Final_2026-08-19.pdf?rlkey=28cy1kdzn2rz0sed173t0thbo&dl=0 Donor Republic is committed to increasing Aboriginal and Torres Strait Islander employment and strongly encourages applications from First Nations people. Candidates must have Australian or New Zealand work rights to be eligible for this position. Contact Brent Frewen at bfrewen@donorrepublic.com.au or 0437 712 707 with any questions about the role.