Consensus
Automated sports prediction aggregator & data science engine.
01 / Overview
An automated sports intelligence platform that aggregates football predictions from multiple online sources into unified, high-confidence verdicts. Created to eliminate tedious manual research across betting tip sites, Consensus autonomously scrapes multi-source predictions and actual match outcomes, enabling quantitative data science, model evaluation, and value-odds detection.
02 / Architecture & Engineering
Built with a Spring Boot backend in Java 21 managing scheduling and fault-tolerant web scrapers to gather predictions across external sources. Raw predictions and historical match results are persisted in a relational store for statistical aggregation and accuracy scoring. SvelteKit provides a snappy, responsive client interface consuming the REST API with fine-grained reactive state updates.
03 / Technical Highlights
Autonomous multi-source web scraper continuously ingesting football match predictions and market odds from leading tipster and analytics engines.
Automated outcome tracking engine capturing final match results to compute empirical source accuracy and eliminate human bias.
Quantitative data science pipeline generating weighted consensus confidence scores and identifying high-value betting opportunities.
High-performance SvelteKit dashboard presenting real-time consensus feeds, odds movement drift, and consolidated match insights in a single view.
Technologies Used
Deliverables
- ▸ Production Dashboard on Vercel
- ▸ Spring Boot Multi-Source Scraper & Data Pipeline
- ▸ Prediction vs. Outcome Tracking Database
- ▸ Statistical Consensus & Calibration Models