Bojan Tunguz launches TabulAI for tabular AI research
Bojan Tunguz, quadruple Kaggle Grandmaster and ex-Nvidia ML engineer, launches TabulAI to advance AI for tabular data. The company targets the research gap in tabular data, which underpins much of business but lags behind images, text, audio, and code. TabulAI emphasizes fundamental research and practical applications. Tunguz, creator of TrainXGB, announced the venture in posts on X.
@tunguz congrats on launching, Bojan!
Today I want to share one of the main projects I have been working on: TabulAI. Tabular data runs much of the business world, yet it has not received the same sustained research attention as images, text, audio, or code. TabulAI exists to change that. 1/9
TabulAI is a research-driven company with a bias toward practical applications — not an ordinary company, and not one I intend to turn into one.
2/9
Today I want to share one of the main projects I have been working on: TabulAI. Tabular data runs much of the business world, yet it has not received the same sustained research attention as images, text, audio, or code. TabulAI exists to change that. 1/9
In practical terms, TabulAI will focus on:
· Basic and applied research in machine learning for tabular data, with special emphasis on classical non-neural-network methods
· Open-source software built from, and alongside, that research
4/9
TabulAI will run research, build open-source tools, engage in education and information sharing effort, and ship practical machine learning products for people and organizations working with structured data. 3/9
TabulAI will run research, build open-source tools, engage in education and information sharing effort, and ship practical machine learning products for people and organizations working with structured data.
3/9
TabulAI is a research-driven company with a bias toward practical applications — not an ordinary company, and not one I intend to turn into one. 2/9
· Free and commercial machine learning products for individuals and enterprises. For example, http://www.trainxgb.com is a fully in-browser GUI for training XGBoost models; we’ll expand substantially on this prototype in the coming weeks.
5/9
In practical terms, TabulAI will focus on: · Basic and applied research in machine learning for tabular data, with special emphasis on classical non-neural-network methods · Open-source software built from, and alongside, that research 4/9
· Educational outreach through lectures, essays, blog posts (such as http://www.xgblog.ai ), social media (@taubl_ai), and other public work
6/9
· Free and commercial machine learning products for individuals and enterprises. For example, http://www.trainxgb.com is a fully in-browser GUI for training XGBoost models; we’ll expand substantially on this prototype in the coming weeks. 5/9
· Educational outreach through lectures, essays, blog posts (such as http://xgblog.ai ), social media ( @tabul_ai ), and other public work 6/9
· Free and commercial machine learning products for individuals and enterprises. For example, http://www.trainxgb.com is a fully in-browser GUI for training XGBoost models; we’ll expand substantially on this prototype in the coming weeks. 5/9
I did the thing.
Today I want to share one of the main projects I have been working on: TabulAI. Tabular data runs much of the business world, yet it has not received the same sustained research attention as images, text, audio, or code. TabulAI exists to change that. 1/9
I did the thing. Go check it out.
Today I want to share one of the main projects I have been working on: TabulAI. Tabular data runs much of the business world, yet it has not received the same sustained research attention as images, text, audio, or code. TabulAI exists to change that. 1/9