We are looking for an experienced ML and backend team to build a reliable football prediction engine for our existing website. This is not a frontend website project. The goal is to create a production-ready backend system that predicts upcoming football matches, starts with the 1X2 market, and later expands to other markets in phases. The system must generate predictions for real upcoming fixtures before kickoff, expose them through an API, and also support JSON/CSV exports for testing and review. In addition to prediction accuracy, the project must include bookmaker odds comparison and value-betting analytics such as implied probability, edge, expected value, ROI, and related reporting. Required core data sources are API-Football, Odds API, and Weather API. Optional advanced data such as xG, player ratings, and team ratings can be added later only if clearly justified and approved. The preferred stack is Python, structured ML models such as XGBoost/LightGBM/CatBoost, FastAPI or Flask, Docker, and a production-suitable database if needed. This project must be delivered in milestones. Milestone 1 is 1X2 only for upcoming fixtures, with working API output, local reproducibility, and clear evaluation. Additional markets and deployment layers will only be added after the first phase is proven and approved. To confirm that you have fully read this brief, please start your proposal with the phrase: Blue Orbit.