Betleads
©2026
Category
Machine Learning · Sports
Location
Bogotá, Colombia
An ML model that reads the Colombian football league.
Predicting the liga colombiana. Finding the odds the bookies underweight.
Betleads is a machine learning system built around the Colombian first-division football league. We feed it historical results, team form, lineups, weather, referee tendencies — the inputs that international bookmakers model but consistently underweight for South American leagues.
The output is a per-match probability distribution we compare against published odds. Where the model disagrees with the market by enough margin, that's a value bet. The system doesn't promise winning every bet — it promises positive expected value over a long sample, with disciplined staking.
Stack: machine learning frameworks in Python (scikit-learn + gradient boosting), live match data piped in through SportsAPI, training and inference running on AWS. The pipeline retrains weekly with new match data. We dogfood the bets ourselves before recommending the product to anyone external.
Credits
- ML model & data pipeline Sebastian Ospina
- Strategy manomano team
Next project
Bolsso
DTC Brand · Bogotá / US
Continue reading →