Applied AI Engineer

We're building a suite of AI products for researchers, and the first is already live: an AI Moderator that runs interviews with participants anywhere in the world, in any language, and turns those conversations into insight reports researchers act on. We're hiring an engineer to own this system end to end. From the interview experience to the synthesis pipeline and the agents we'll build next, you'll continuously improve a product that directly impacts customer experience and business outcomes. 🌍 About Respondent Respondent is building the infrastructure that powers participant recruiting for the research industry. Our mission is simple but ambitious: to make it effortless for researchers to find high-quality human participants - whether they’re testing a product, exploring a market, or validating an idea. From our founder’s first recruits on a NYC subway to a global panel of 4 million participants across 150+ countries, we now support research for companies like Stripe, Microsoft, and Canva. Our panel is one of the most diverse and responsive in the industry. We’re a 30 person team across Sydney, Manila, and the US. We’re intentional about collaborating in person - serendipitous hallway conversations, whiteboarding sessions, and quick alignment cycles help us move faster and build something great together. We’ve gone from 0 to 1, recently reaching profitability and intending to stay so. We’re scaling with intention and reinvesting in the business to grow our platform, panel, and team. Joining at this stage means your work will have a direct impact as we take Respondent from 1 to 10. 💼 The role: You'll be joining our Sydney team as a Senior Applied AI engineer working closely with Product and Engineering to own and evolve the AI Moderator, and to build the agents that come after it. This is an in-office role, and we’re looking for someone who’s genuinely excited about collaborating in person. Day to day, you'll raise the quality bar on the moderator itself - conversation quality, multi-language support, the voice pipeline - and on the synthesis engine behind our insight reports. You'll work across rich interview data (transcripts, recordings, screen captures), apply AI thoughtfully where it adds real value, and partner closely with Product to define clear success metrics and move them. There are interesting challenges in how researchers interact with their data: resurfacing key findings, querying across studies, and making our reports the best in the industry. This is commercially grounded work - shipping improvements quickly, measuring impact, and iterating. 📌 What You’ll Do Own and evolve the AI Moderator: Improve how it conducts interviews - question quality, adaptivity, multi-language support - so researchers get great studies quickly, anywhere in the worldMake our insight reports top notch: Design and improve the synthesis pipeline that turns raw interviews into reports researchers trust, and build ways for them to interact with their data and resurface key findingsBuild the product across the stack: Architect APIs and services that handle real-time conversation, media, and large volumes of interview data - and build the researcher-facing experiences on top of themBuild robust evaluation and observability: Define quality metrics, build automated evals, and instrument production systems so we can measure regressions and improvements with confidenceOptimize cost and quality of generations: Profile token usage and experiment with model routing, prompt engineering, and caching to hit the right quality, latency, and cost tradeoffs at scaleShip AI features end-to-end: Work closely with Product to turn AI capabilities into reliable, production-ready features - not just prototypes 💡 What You’ll Bring Strong full stack foundations (Typescript first): Deep experience building scalable services and APIs, plus the frontend chops to ship the product on top of them. Comfortable with large datasets and reasoning about performance. Our stack includes TypeScript, Node.js, React, AWS and MongoDB - but we care more about your ability to ship than a perfect tech match.Deep LLM/AI engineering experience: You've built and shipped LLM-powered features in production - conversational systems, agents, or synthesis - and you understand prompt engineering, retrieval, embeddings, and the practical tradeoffs between approaches. Fine-tuning experience is a plus.Data and evaluation mindset: You know that AI quality starts with data quality. You've built evaluation datasets, designed annotation guidelines, or constructed eval pipelines. You think in metrics - quality, faithfulness, cost-per-query - not vibes.Builder mindset: You've taken an AI system from "it kind of works in a notebook" to "it runs reliably in production and we can measure it." You move quickly and turn messy, ambiguous problems into working systems.High ownership: You take a feature from problem to shipped and keep it moving on your own - you share progress proactively, raise blockers early, and ask for help when you need. You define problems, run experiments, ship, measure, and iterate without heavy process. 🤝 Why Join Us? Real impact: You'll be part of a small, high-ownership engineering team where your work ships fast and directly shapes the experience for customers like Stripe and Canva.Profitable and growing: You'll get the excitement of startup growth with the stability of proven traction and profitability.Modern tech and mindset: You'll work with the latest tools and AI-powered dev practices, giving you a front-row seat to cutting-edge engineering.Culture that cares: You'll join a tight-knit global team (Sydney, Manila, US). We're collaborative, supportive, and believe in balancing high performance with a fun workplace.