I’m building a subscription-based web app that lets a Registered User log in, spend credits, and receive high-quality text generated through large language models. The core workflow will call OpenAI or OpenRouter for the text work, and I’ll lean on Replicate’s API later for image or video add-ons, so the architecture has to be modular. Here’s the flow I need implemented end-to-end: • FastAPI backend that exposes a secure REST/GraphQL layer, routes traffic to the chosen LLM, stores usage counts, and publishes errors to Sentry. • React or Next.js frontend that handles sign-up / sign-in, credit balance display, a prompt editor, and the results viewer. Only the Registered User role is required right now; an admin interface can wait. • Stripe integration for monthly subscriptions or one-off credit packs, with customer portal, webhooks, and proper metered usage tracking. • Sentry wired into both client and server so I get real-time alerts on exceptions. • Clean repo with environment-based settings, Docker dev setup, and clear README so I can deploy to my own cloud later. Required Skills: - Large Language Model workflows (OpenAI / OpenRouter) - Replicate API (image/video generation) - Stripe (subscriptions, payments, customer portal) - Sentry (error tracking for backend + frontend) - Web application development (FastAPI + React or Next.js) If you have shipped something similar with FastAPI, React/Next, Stripe, and Sentry, let’s talk details and timelines.