Gamified Pseudo-Auction platform build required

We are launching cashforsmartphones.com, a high-conversion phone buyback aggregation platform targeting the UK and Irish markets.
This is NOT a standard, boring table-comparison site. Instead, it is built as a Pseudo-Live Flash Auction. The consumer inputs their smartphone model and condition, triggering a 15-second live interactive "bidding war" animation between major recyclers (MusicMagpie, Envirofone, Mazuma, etc.). The system calculates the true highest scraped market value on the backend and displays it dynamically as the winning bid. When the user clicks "Accept Bid," they are transferred via a custom dynamic affiliate deep-link to the target reseller to finalize the payout.
The front-end design must be dark-themed, ultra-premium, high-converting, and heavily optimized for mobile taps (resembling a fast-paced fintech app or stock-market ticker).
Core Tech Stack
• Frontend: Next.js (React), Tailwind CSS, Framer Motion (for the real-time cascading bid animations).
• Backend: FastAPI (Python) or Node.js to manage parallel asynchronous web requests.
• Database/Caching: Redis (or simple in-memory cache layer) to store daily fetched phone price metrics.
Core Functional Requirements
1. Multi-Step Conversational Onboarding (Frontend)
• Clean, zero-friction 4-step selector form: Brand ➔ Model ➔ Capacity ➔ Condition (Pristine, Good, Poor, Faulty).
• Auto-suggestions or simple dropdowns for quick access to core models (e.g., iPhone 13 through iPhone 17 Pro Max).
2. The 15-Second Flash Auction Engine (The Core Feature)
• Once submitted, transition to a custom loading page that behaves like a live auction floor.
• Frontend State Machine Animation: Bids must pop up consecutively every 600ms–800ms with a rising value counter.
• The Pacing Logic: The backend will provide the true highest and lowest scraped prices. The frontend will generate 5–6 transitional steps (starting at ~70% value) assigning fake rising counter bids to partner brand logos, concluding exactly at the real maximum scraped price for that device tier.
3. High-Speed Asynchronous Pricing Backend
• Parallel API execution or web crawlers to poll internal XHR endpoints of the top 5 recyclers simultaneously when a model is queried.
• Implement a 24-hour cache layer (Redis or lightweight server-side storage) so that if a device model has already been searched that day, the system serves the values instantly (<150ms) instead of running a fresh scrape.
4. Transition Intermediary Page & Affiliate Deep-Linking
• When "Accept Bid" is selected, trigger a 3-second animated intermediary screen showing confirmation text: “Securing £XXX Offer Token... Transferring to [Partner] Portal.”
• Dynamic formulation of target affiliate links using deep-link paths for Awin/Impact networks.
Deliverables & Milestones
• Milestone 1 (UI/UX Implementation): Interactive multi-step selector setup and complete 15-second Framer Motion animation loop using mock JSON datasets.
• Milestone 2 (Backend Core): Python FastAPI script or Node service mapping the asynchronous lookup framework across the top 4 target platforms, with a simple cache management layer.
• Milestone 3 (Integration & Deployment): End-to-end integration of frontend events with real data feeds, configuring the intermediary transfer screen, and launching on Vercel/Render.

Applicant Requirements
• Strong portfolio showcasing fluid React web animations via Framer Motion or GSAP.
• Clear experience with web scraping using Playwright / Puppeteer / Beautiful Soup or manipulating internal API requests.
• Familiarity with setting up affiliate tracking links or deep-linking architectures is a major plus.
Please provide examples of previous interactive dashboards, gamified web components, or scraper engines you have engineered when applying.