PORTFOLIO

THE TEXAS A&M FOOTBALL HALL OF FAME

Journey through Aggie football history with a single click. Hear their stories. Celebrate the longstanding tradition of good old fashioned traditional traditions. Take pride in feeling proud of your pride. Visit THE TEXAS A&M FOOTBALL HALL OF FAME.

OpenAI GPT-4.1_nano DALL-E 3 Convex Vite React TypeScript Tailwind Three.js React Three Fiber Stripe Workers


OVERVIEW

Step into the hallowed halls of the most storied football program in College Station, TX. Click the TELL ME A STORY button and learn about the heroic exploits of an Aggie football legend. Every visitor receives a totally unique story, a “Twelfth Man Rating,” and a bespoke, AI-generated image of their Aggie football legend. All served in realtime.

EXPERIENCE

  • TELL ME A STORY: One click procedurally generates a unique player's biography line‑by‑line.
  • TWELFTH MAN RATING: When the story ends, the player receives a “Twelfth Man Rating” via a 3D visualizer that renders the rating as a series of zero to twelve 3D models of football players.
  • BESPOKE IMAGE: The app extracts key details from the biography, adds a random activity via unseen backend logic, concatenates the results, and generates a 1024×1024 image via OpenAI (DALL·E 3).
  • DONATE: A Stripe‑powered tip jar offers various donation options. Completed checkouts show a thank‑you screen and return users to the experience.

HOW IT WORKS

  • FRONTEND: React 19 + TypeScript + Vite, styled with Tailwind.
  • 3D: The Twelfth Man Rating visualizer and other 3D assets were built using Blender and are rendered with Three.js via React Three Fiber and Drei with guarded fallbacks.
  • BACKEND: Convex for data, queries / actions, and storage; handles structured randomization, image orchestration, and retrieval.
  • AI: GPT-4.1_nano & (DALL·E 3) for generating bespoke prompts & images. Prompts are engineered and sanitized to preserve uniform fidelity and safety.
  • PAYMENTS: Stripe Checkout with signed webhooks; a small Express service verifies signatures, rate‑limits the endpoint, and forwards events for finalization.
  • RELIABILITY: Error boundaries for 3D visualizer, prompt sanitization, base64 image handling, and careful sequencing of the story → rating → image UX flow.

WHAT MAKES IT DIFFERENT

  • Hybrid Creativity Model: Stories use deterministic, structured randomization — fast, coherent, and low-cost — to ensure reliability. API calls are reserved for generative tasks where they matter most.
  • Seamless Flow: One click advances story → rating → image automatically — no fiddly controls.
  • Brand Fidelity: Image prompts and biography randomization engine copy always enforces Texas A&M themes, colors, and other identifiable school characteristics for ultra-specific identity.
  • Embodied Humor in 3D: The Twelfth Man Rating isn’t just text — it’s a living formation with emissive maroon highlights, subtle rotation, and paced timing to help the effect land reliably.
  • Production Polish: Prompt sanitization, error boundaries, base64 handling, signed Stripe webhooks, and endpoint rate limiting.

MY ROLE

  • Concept, interaction design, narrative constraints, and copywriting
  • Frontend engineering (React, R3F), 3D integration, and hall of fame UX design
  • Backend functions, real‑time data model, randomization engine, prompt engineering & sanitization, and Stripe integration
  • Performance tuning and type‑safe refactors across the codebase

OUTCOME

A fast, delightful, and shareable fan experience that blends tradition with modern web tech—engineered for reliability today and room to grow (social sharing, galleries, auth, analytics).

TECH STACK

  • React 19, TypeScript, Vite, Tailwind CSS
  • Three.js + React Three Fiber + Drei
  • Convex (Real-Time DB / Serverless Functions / Anonymous Auth)
  • OpenAI Responses API (GPT-4.1_nano) + Images API (DALL·E 3)
  • Stripe Checkout + Webhooks
  • Cloudflare Workers