Kaggle launched in 2010 as a platform for machine-learning competitions: an organization posts a dataset and a scoring metric, and anyone in the world can submit predictions and climb a public leaderboard. It was founded by Anthony Goldbloom (with co-founder Ben Hamner). Kaggle’s own about and history pages were not reachable at their expected URLs during verification (HTTP 404), so the precise founding date here is corroborated by secondary coverage rather than a live primary founding document - see the FLAGGED note in the accompanying report.
The competition model made machine learning a measurable, repeatable discipline. Instead of arguing about which method was best in the abstract, teams competed on identical data with identical metrics, and the leaderboard settled it. This produced reusable techniques (gradient-boosted trees and large model ensembles became competition staples) and surfaced talent: strong Kaggle finishes became a recognized credential, and the platform served as a hiring funnel for data science teams.
Kaggle today describes itself on its homepage as “The World’s AI Proving Ground.” It also grew into a widely used host for public datasets, shared notebooks, and pretrained models. Google acquired Kaggle in 2017, after which it continued operating as a community platform. Its lasting contribution to the open-source and frameworks story is cultural: it normalized open benchmarks and shared, runnable code as the way the field measures progress.