1. Our paper was accepted by ICML 23, which is named 'Trompt: Towards a Better Deep Neural Network for Tabular Data'. We propose a new deep learning model architecture to better deal with tabular data, it is the state-of-the-art model in the deep learning models and is competitive with the GBDT-based models.2. Responsible for generative AI related projects and technology research and development.3. Build the AutoML to complete the precision marketing of various products in different departments, which is 2~3 times higher than the expert method.4. Carry out technology research and development, and study the related applications of external labels.

Data analysis and artificial intelligence research.

Use deep learning to build the nlp platform and complete a number of tasks in different departments.

Use the online behavior data of the customer to complete the application of the recommendation mechanism.

Responsible for data exploration and analysis.