Oracle Labs Presentation & Tech Talk

Research in Oracle Labs: AI, LLM, Vector Database, Data Analytics and Business Applications

This talk focuses on Oracle Labs’ role in the rapidly changing world of AI and data analytics. We will showcase Oracle’s latest research and developments in these areas, emphasizing how AI is transforming business practices.

We start with an overview of how database is evolving for AI workload. We introduce several vector database features announced in the recent Oracle Cloud World and discuss technical challenges for academia to solve. We show native support of graph data modelling in database and its large-scale, elastic execution, which enables to manage complex data relationships and extract insights from huge datasets. We’ll discuss the technical hurdles and recent advancements in scalable data processing.

Next, we discuss Oracle Lab’s work in applying AI to enterprise applications. This includes an overview of how Oracle Lab’s Recommendation System is improving user experience and business intelligence for enterprise applications. We also cover how Oracle’s AutoMLx toolset incorporates fairness metrics, helping users to add considerations into their AI applications. We then show how modern AI is crucial for addressing issues in cybersecurity as well.

Lastly, we explore other uses of Large Language Models (LLMs) in solving business problems, with a focus on Code Generation and Data Integration. We’ll discuss the challenges of using LLMs for practical business applications, particularly in code generation, and highlight our approaches. In Data Integration, we’ll examine how LLMs help in merging different unstructured data sources.

Sungpack Hong is a Senior Research Director at Oracle Labs. He joined Oracle at 2012 after getting his PhD from Stanford University. He leads several research projects regarding graph data processing, large scale data analytics, domain-specific languages, and machine learning.