Year of the project: 2016
Automated system for bookmakers that provides facility for betting on victory of team or in-play events in e-sports (games like Dota 2, League of Legends, etc.).

Primary users of the system are bettors, who place bets on website.

What I have done during this project:
- Created system's architecture and implemented its core components;
- Collected massive datasets of Dota 2 professional matches snapshots (~4 million records), leagues, players, etc. from DotaBuff and Steam API;
- Created statistical models to predict three Dota 2 in-game events ("What time will be first blood", "Will Roshan be slain during the game", "What tower will be destroyed first");
- Implemented odds generation using statistical models results;
- Implemented website, which displayed data analysis results and current state of a match (matches tracker for Dota 2 professional league matches like on http://www.trackdota.com).

The work was done for BetGame startup.