AI Full-Stack Engineer

DESCRIPTION We're a fashion discovery product with three shipping surfaces and one in flight: a server-rendered React web app on Cloudflare Workers, a React Native app built with Expo, a Postgres backend on Supabase, and a Shopify embedded app that puts our enhanced AI shopping assistant on a merchant's own storefront. Behind all four sits a scheduled pipeline that acquires and normalizes structured product data from every brand in the catalog. We need someone who can move across all of it without a handoff and who builds with the next order of magnitude and scaling in mind. Our catalog is on track to grow several times over in the next year, and both the surfaces on top of it and the pipeline underneath need to work for existing customers and also scale for future customers with growing user base and application interactions. This is a wide role. We are explicit about that because the wrong candidate for it is a specialist who wanted one lane, and the right one is somebody who can wear multiple hats. WHAT YOU'D BUILD Ship the mobile app, then keep it level with the web app. It is built and verified end to end (swipe feed, conversational agent, push notifications, onboarding); what is left is TestFlight, publishing it to the App Store, and planning out the Android development. Product Frontend and Backend: Improve the existing web and mobile application: make the product discovery page more robust, improve chatbot functionalities with more tools, decrease latency on the chatbot, improve prompting, and make sure the security vulnerabilities are limited. Develop additional features for the web and mobile application with clear requirements from technical co-founder. Scaling the application itself—scaling our database for more brands, catalog, appropriate data and scaling our product for more users so it can be user friendly. Scaling the pipeline underneath it, from dozens of sources twice weekly to several hundred daily. The architecture is specced and being built alongside you; you own making it real and keeping it healthy: per-source config and rate limits, egress routing, polite concurrency, retry and escalation when a source starts refusing traffic, and monitoring that catches a source changing shape before bad data reaches users. This half is operational and unglamorous, and it decides whether the product has anything to show. THE STACK React 19 with TanStack Start and TanStack Router, server-rendered on Cloudflare Workers. TanStack Query for server state. Tailwind 4 with Radix and shadcn/ui. Expo SDK 57 / React Native 0.86 with expo-router. Supabase for auth and data, accessed through Postgres functions rather than raw queries. Shopify’s React Router app framework, App Bridge, Polaris, theme app extensions. TypeScript strict throughout. Bun. On the pipeline side: TypeScript on Bun, headless browser automation for the sources that need it, and Postgres. Shopify’s Admin GraphQL API and bulk operations for authenticated brand syncs. We are planning to migrate appropriate portions of our codebase to Cloud for scaling purposes so Cloud knowledge is expected for this role, preferably Google Cloud Platform, to assist with planning and some implementation if required. WHAT WE'RE LOOKING FOR Strong React and TypeScript. Strict mode. You are comfortable with server-side rendering and understand why an SSR bundle boundary matters. Real React Native shipped to at least one app store. SQL comfort. Our data access is through Postgres functions with row-level security. You do not need to write the hard ones, but you need to read them and know when the answer is a schema change rather than a client-side workaround. Taste and great sense of UI. This is a fashion product; the interface is a large part of the value. We are not looking for a bootstrap-default aesthetic. You have taken an application through real growth and can name what broke. You measure the front end. You have used real-user metrics rather than a local Lighthouse score, and you have an opinion about what a performance budget should be enforced against. Hands-on experience running high-volume data acquisition in production where proxies and egress routing, per-source rate limiting, concurrency control, partial failure, and retry-and-escalation strategy were involved in the development process. Working knowledge of Google Cloud or any other cloud platforms that can be translated to another cloud platform.