Panels

Panel A

Data + AI: Enhancing the Database Community's Impact on Data-Centric AI Systems

Panel moderator: Guoliang Li
Panelist: Amr El Abbadi, Lei Chen, Gao Cong, Feifei Li

Abstract: Large Language Models (LLMs) have revolutionized numerous applications, and many of their techniques are intricately connected to our database community. It is essential for our community to play a pivotal role in data-centric AI systems, which encompasses areas such as data preparation, data cleaning, data integration, data selection for LLM, LLM training systems, LLM inference systems, LLM for database systems, and LLM for multi-modal data management.

In this panel, we will explore the following key questions:

1. How can our database community contribute effectively to AI systems?

2. In what ways are LLMs revolutionizing database systems?

3. How do we construct robust Data + AI systems?

Guoliang LI

Amr EI Abbadi

Lei Chen

Gao Cong

Feifei Li

Industry Panel Session

What is the Role of Database Systems for Generative AI Applications?

Panel moderator: Reynold Cheng (The University of Hong Kong)
Panelist: Wenjie Zhang (University of New South Wales), Bo Guo (Baidu), Rong Zhu (Alibaba), Jianjun Chen (ByteDance), Jinwei Zhu (Huawei)

Abstract: Each panelist gives 10-minute presentation on sharing their uses of database systems as the backbone for effective, scalable, and context-aware generative AI applications.

In this panel, we will explore the following key questions:

1. How do vector databases support semantic search and retrieval-augmented generation (RAG)?

2. Does text-2-SQL play a role in GAI?

3. How do data pipelines and data preprocessing impact GenAI performance?

4. How do database architecture choices impact the scalability and efficiency of GAI?

5. How do different kinds of databases (e.g., relational, document, and graph databases) contribute to generative AI workflows?

6. How to ensure data quality, privacy, and compliance in databases that support GenAI workflows?

7. How to support AI models that work with text, images, audio, and video through vector and multimodal databases?

Reynold Cheng

Wenjie Zhang

Rong Zhu

Jianjun Chen

41st IEEE International Conference on Data Engineering, Hong Kong SAR, China – May 19-23, 2025