Back to HBA

FameForecast

A live streaming discovery platform built from zero — multi-platform creator integrations, real-time competition and prediction systems, payment infrastructure, AI-assisted content tooling, and full deployment infrastructure. Built entirely by Michael Hawkins.

100%
Founder-built, no dev team
10+
API integrations
3
Payment rails live (PayPal, Square, Crypto)
Live
Production deployment, operating

What is FameForecast?

FameForecast is a live streaming discovery platform where fans predict creator growth, compete in seasonal competitions, and collect creator cards. It aggregates data across Twitch, YouTube, Kick, Instagram, Spotify and TikTok — all built on a custom Flask backend with a full-featured admin control plane.

Every layer — from database schema design to payment webhooks to email automation to the real-time scoring engine — was architected and built by Michael Hawkins as both a commercial product and a proof-of-concept for the systems HBA brings to client engagements.

The platform is live in production, processing real payments, sending real emails, and running real competitions.

Tech Stack

Python 3.13 Flask SQLAlchemy Flask-Migrate (Alembic) PostgreSQL Redis (caching) Google Cloud Storage Google App Engine Flask-Login Flask-Mail Authlib (OAuth 2.0) Flask-WTF / WTForms Flask-Compress Flask-Caching Jinja2 Bootstrap 5 Twitch Helix API YouTube Data API v3 Kick API Instagram Graph API Spotify Web API TikTok API PayPal REST API Square Payments API CoinSpot (Crypto) Apify (Residential Proxies)

Capability Areas Demonstrated

Creator Platform Systems

  • Twitch, YouTube, Kick, Instagram, Spotify and TikTok API integrations
  • Creator profile and identity pipeline with enrichment
  • Search aggregation across all platforms in a single interface
  • Creator application and onboarding workflows
  • Partner creator management and showcase tooling

Commerce & Payment Infrastructure

  • PayPal REST API — purchase flows, webhooks, confirmation
  • Crypto payment rail via CoinSpot with settlement verification
  • Subscription purchase and premium tier management
  • Checkout session continuity across on-site → provider → return
  • Transaction logging and purchase history

Real-Time Systems & Runtime Delivery

  • Real-time scoring engine for competitions and predictions
  • OAuth 2.0 with token lifecycle management via Authlib
  • Redis caching layer for API responses and session data
  • Background task scheduler for data refresh and settlement
  • Telemetry validation and interaction integrity controls

Data Engineering & Governance

  • Multi-model relational schema with Alembic migration management
  • Creator and artist data persistence with enrichment pipelines
  • Seasonal scoring with materialised views and aggregates
  • Virality scoring algorithm across platform metrics
  • Category taxonomy management with bulk tooling

Applied AI Workflow Systems

  • AI-assisted card generation pipeline for creator collectibles
  • Structured profile authoring using local model inference
  • Content safety and deduplication controls
  • Admin review gates before production writes

Email Automation & Outreach

  • Flask-Mail transactional email — welcome, verify, reset, updates
  • Creator invitation and bid notification emails
  • Subscription and competition result notification flows
  • Email open tracking via pixel

Architecture Highlights

  • Blueprint-modular Flask app

    15+ blueprints covering auth, payments, predictions, competitions, creators, admin, showdowns, and resources — all independently testable and deployable.

  • Security-first implementation

    CSRF protection on all forms, DOMPurify XSS sanitisation, bot-trap routes, a proprietary bot-resistant human signal scoring engine for competition integrity, and input controls throughout.

  • Performance engineering

    GZip response compression, parameter-inclusive cache keys (MD5), tiered invalidation, lazy-loaded assets, PurgeCSS-generated per-page stylesheets, and preconnect for all external origins.

  • Cloud-native deployment

    Google App Engine with App Engine warmup, Redis on Cloud Memorystore, Google Cloud Storage for user-uploaded media and card assets, structured JSON logging throughout.

  • Dual-build segmentation

    Public and R&D builds maintained simultaneously. R&D features gated behind runtime controls and not exposed in the public deployment.

  • Proprietary video classification engine

    A custom NLP-based pipeline for classifying and continuously re-categorising Clip Duel videos at scale — built entirely in-house.

What this demonstrates for your engagement

FameForecast is not a side project or prototype. It is a live, revenue-generating platform built from first principles — every system was scoped, designed, and implemented end-to-end.

When HBA takes on a client engagement, this is the standard of delivery you can expect: production-grade systems, full traceability, no black boxes, and architecture that your team can understand, maintain, and build on.

Start a conversation Connect on LinkedIn Visit FameForecast