ABOUT REFINE A spreadsheet error in one of the most influential economics papers of the decade helped justify austerity across Europe and the US. Millions lived with the consequences. The error was only found years later, by sheer luck, long after the damage was done. Flawed analysis shapes policy, investment and lending every day, costs billions every year, and mostly goes unnoticed. AI makes it worse: rigorous-looking work is now cheap to produce. Refine exists to catch these errors before anyone acts on them. We're building the infrastructure that verifies high-stakes work; think CI/CD for everything beyond code. It reviews a document like a team of experts, checking the maths, data, logic and argument, in minutes. - The best harness. Refine 5 beats frontier models such as GPT-6 Astra in over 90% of blind head-to-heads across ten fields, from economics to biology. It combines the top frontier models at hundreds of times normal compute, so no one model's blind spots get through. We tune it on documents seeded with the errors frontier models miss. - The de facto AI reviewer for economics. Econometrica and the AER run every paper they publish through Refine. It's used at 24 central banks and by researchers at Oxford, MIT, Stanford, Yale and Cambridge; Cambridge, UCLA and NYU provide it to theirs. - Beyond academia. 23,000+ reviews sold at 1,000+ institutions, including companies checking filings, tenders, compliance codes and technical specs. - Enterprise-grade. ISO 27001 certified, SOC 2 Type 1 (Type 2 in progress). On the web and inside ChatGPT, Claude and Codex. THE ROLE You'll own growth end to end, from first visit to renewed subscription, building it from the ground up with room to build a team later. Priority: top of funnel and fast experimentation. You're measured on signups, active, paying and repeat users, and on revenue as we grow from $1M to $10M. You'll work directly with the founders, Yann Calvo Lopez (CEO) and Ben Golub (Chief Scientist, formerly a tenured economics professor at Harvard). Our advisors include OpenAI co-founder John Schulman. WHAT YOU'LL DO - Experiment constantly. Tests on pages, offers, onboarding and pricing. - Open the AI channels. ChatGPT ads, and being the answer when someone asks an AI how to check a paper or verify a contract. Turn our ChatGPT, Claude and Codex integrations into distribution. - Win search, classic and AI. Drive search traffic and get cited in AI answers. - Make paid acquisition pay off. - Grow the viral loop. Users share reviews with co-authors and colleagues, who sign up to read them; some become paying users. - Build lifecycle email around the deadlines that drive usage: submissions, grant cycles, filings, tenders. - Whatever else makes us grow. If it moves the numbers, it's in scope. OUR DATA (YES, REALLY) Everything is in PostHog and wired to AI: every product event captured server-side (uploads, reviews, purchases, credits); Stripe, Clerk, Customer.io http://Customer.io, ad platforms and our app database in the same warehouse; plus model costs, logs, feature flags and experiments. Ad click to revenue is one query, per person or per organisation. Ask an AI for "signup-to-paid by campaign, last 60 days" and set up the next test in the same conversation. No begging engineers for data. WHO YOU ARE - A hands-on operator with about 4-7 years in growth at a self-serve (B2C or prosumer) or B2B SaaS company. Less is fine if you're great. - In the data daily. You ideally don't wait for someone to build you a dashboard. - AI-native. You use Claude, Claude Cowork, Codex or similar every day to research, analyse, build and test, and can show examples. - An experimenter. Five small tests this week beat one big plan next month. - Proven in paid acquisition (Google Ads at minimum), measured in revenue, not clicks. - A clear writer. We use AI for almost everything except customer-facing copy: you write that, for PhDs, law firm partners and central bank economists. - An owner. Idea to measured result, low ego, comfortable with ambiguity. Nice to have: PostHog or similar analytics; setting up lifecycle email (Customer.io http://Customer.io or similar); coding agents (Claude Code, Codex) to build pages and automations yourself; LinkedIn Ads; range across B2C and B2B; Spanish or another European language. HOW WE WORK - AI-native throughout. We build with AI and expect marketing to as well: one person with the right setup does what used to take a team. - Autonomy. You choose what to work on, by impact. - Speed. Tracking fixes, pages and product tweaks usually ship the same day. - Evidence over opinions. If we can't measure it, we don't claim it worked. - Remote-first. Async by default, few meetings. - Respect for users. They're sceptical experts; we win them with substance, not hype. - Meritocracy. Talent, hard work and results are rewarded. Compensation: competitive. HOW TO APPLY Look around refine.ink http://refine.ink, then send your CV or LinkedIn with short answers to: 1. What's the first thing you'd do if you joined Refine? 2. How do you use AI in your marketing work? Show us something real: a workflow, a prompt, an automation. How you think matters more than polish.