Oddswell

Overview

Oddswell is a mobile app for AI-assisted sports predictions. Pick a sport, fill a matchup form, and get structured analysis with confidence scores, odds context, and risk flags. Built with Expo SDK 56, expo-router, and OpenAI server routes, each analysis runs through a server-side pipeline that combines live web research, structured JSON output, and numeric engines where applicable.

Analyses save to local history on the device (SQLite via Zustand persist). No account required. Other sports on the home grid (tennis, basketball, and more) are placeholders with no pipeline yet.

Supported Sports

Live analysis pipelines:

• MMA - Fighter matchup form, structured predictions blended with market and strength math
• Boxing - Same combat pipeline as MMA with sport-specific UI
• Football (1X2) - Probabilities from a numeric engine (market + Poisson); the LLM researches and explains, it does not own the headline numbers
• FIFA World Cup 2026 - Tournament winner predictions with dedicated screens

Key Features

This template includes:

• Server-side AI pipeline - Web research, structured JSON schemas, and sport-specific numeric engines
• Three analysis tiers - Fast, Standard (default), and High. Users choose in About → Analysis
• Sport-specific result UI - Tabs, confidence, odds context, and risk flags per sport
• On-device history - Zustand + expo-sqlite/kv-store. Analyses persist locally without a backend
• Expo API routes - POST /api/analysis/mma, /box, /futbol, and /worldcup-winner
• Native navigation - iOS form sheets for new analysis, full-screen loading and result cards, tabs hidden during analysis flow
• i18n ready - i18next wired for multi-language support

User Flow

Home → Sport hub → New analysis (sheet) → Loading → Result → History

1. User picks a sport and fills a matchup form.
2. A loading screen calls the server-side analysis pipeline.
3. Results render in sport-specific UI with tabs, confidence, and risk flags.
4. Analyses save to local history for quick recall.

Analysis Quality

Users choose Fast / Standard / High in About → Analysis. Default is Standard.

• Research - live web context for up-to-date matchup data
• Structured - JSON schema predictions for consistent, parseable results

Tech Stack

• Framework: Expo SDK 56 (React Native 0.85, React 19)
• Navigation: expo-router with tab and stack layouts
• Language: TypeScript
• AI: OpenAI via Expo API Routes (app/api/analysis/*)
• State: Zustand + expo-sqlite/kv-store for analysis, history, and tier preferences
• i18n: i18next
• Auth: Optional EXPO_PUBLIC_API_SECRET header for client → API calls

Real-World Application

Sports analysis technology powers multi-billion dollar industries including betting platforms like DraftKings, media companies like ESPN, and analytics firms serving professional organizations. Oddswell provides the foundation for building similar applications with modern React Native and AI capabilities.

The app processes fighter statistics, historical data, and real-time odds to generate intelligent predictions that rival professional sports analysts. A real-world test case was UFC 318, where the AI achieved 9 out of 14 correct predictions, demonstrating professional-grade accuracy. See the full prediction results for detailed analysis.

What You Get with Pro Lifetime Access

When you become a Pro member at Code with Beto, you unlock lifetime access to all our premium resources. Including: complete source code, templates, and tools through our private GitHub organization. Every project we release is available there, searchable and ready to use.

We actively maintain everything to ensure a smooth dev experience across iOS and Android, with easy setup and up-to-date dependencies. Pro access also includes entry to our private Discord community, where you can ask questions, get support, and connect with other developers.

Get Started Today

Questions? Join our Discord server where our community of developers is ready to help! Or if you prefer, send us an email at beto@codewithbeto.dev