NVIDIA's Kaggle Grandmasters team achieved first place in the 2026 NeuroGolf Championship and second place in the KDD Cup 2026 Data Agents Competition. In NeuroGolf, they used manager agents to coordinate specialized workers and build a shared knowledge base, while for KDD, a purpose-built harness narrowed the performance gap between a frontier LLM and a smaller open model.
SynthePulse Insight · AI deep reading
Version 1 · 1 source
NVIDIA's AI team won first place in the 2026 NeuroGolf Championship and second place in the KDD Cup 2026 Data Agent Competition. The former relied on a manager agent coordinating specialized workers and building a shared knowledge base, while the latter used a customized framework to narrow the performance gap between frontier large models and smaller open-source models.
On July 24, 2026, the official NVIDIA AI account on X announced that its Kaggle Masters team achieved outstanding results in two major competitions: first place in the 2026 NeuroGolf Championship and second place in the KDD Cup 2026 Data Agent Competition. Team members include Chris Deotte, Eduardo Rocha de Andrade, Gilberto Titericz Junior, Jiwei Liu, and Maximilian Jeblick.
In the NeuroGolf Championship, the NVIDIA team employed a manager agent to coordinate multiple specialized worker agents and built a shared knowledge base containing reusable tools. This architecture enabled efficient collaboration, ultimately securing the championship.
In the KDD Cup Data Agent Competition, the NVIDIA team developed a customized framework that successfully narrowed the performance gap between frontier large language models and smaller open-source models. This achievement demonstrates that with a well-designed framework, smaller models can approach the performance of larger models on specific tasks.
This report is based on a tweet from the official NVIDIA AI X account, a first-party official source, but does not provide detailed technical reports or third-party verification.
Through agent coordination and a customized framework, the NVIDIA AI team demonstrated its technical prowess in two top-tier data science competitions, providing practical examples of agent collaboration and model optimization.
Primary report
Primary source