Cloud Coaching - GenAI in OCI Services: Observability and Management & Autonomous Database

Introduction

Introduction

Good morning everyone and thanks for joining us. Hope you're all feeling well. We're still waiting for a few more people to join, so just keep an eye on the counter as we have a pretty large group today.

Let's get started as you're more interested in the man of the moment, Herman, a longtime presenter here at Oracle, presenting some of our biggest sessions. Herman, over to you.

Session Beginning

Thank you, Thomas, for that great intro. Let me enable my camera and share my screen.

Hello everybody, get ready for an interesting session. Today, we will cover AI-related sessions and how employing Generative AI (GenAI) is enhancing Oracle services, specifically focusing on the Oracle Autonomous Health Framework, Observability and Management Platform, and finally, Select AI.

Data Management and AI

Data is crucial in business operations, and more so with AI advancements. AI is now embedded at every layer of the application stack, making data management more efficient. Oracle is leveraging AI to improve our internal tools and services to make it easier for customers to use our services.

Implementation

• AI in Infrastructure: Incorporating AI in Exadata infrastructure and software.
• Enterprise Data Management: AI in Oracle Database 23c, supporting AI-based workloads.
• Application Stack: Embedding AI for transforming productivity.
• Service Layer: AI services for specific needs like anomaly detection, digital assistance, and speech and language processing.

Autonomous Health Framework (AHF)

The AHF provides a diagnostic assistant that can answer detailed questions using both APIs and large language models (LLM). Whether it's diagnosing log writer issues, understanding node evictions, or general queries about database performance, the AHF assistant simplifies these tasks.

Use Cases

• Intent Classification: Understanding what users are trying to do.
• Entity Extraction: Identifying the entities in user queries.

Examples

• Query: "What is the purpose of log writer process?"
  • Response: Detailed answer based on given entities.
• Query: "Are there any OS issues identified 15 minutes before 3am?"
  • Response: Visual reports.

Observability and Management Platform

This platform focuses on monitoring the infrastructure, including health and workload monitoring. It utilizes an AI-based conversational interface for better insights and problem diagnosis.

Examples

• Query: "Why is my database slow?"
  • Response: Details about what's happening with suggestion.

Conversational Insights

The conversational insights can answer complex questions like “What are the top processes utilizing most GPU memory in the last six months?” which make life easier for breach engineers and Ops teams.

Introducing Select AI

Select AI allows users to query their databases in natural language. By creating an AI profile and connecting to an LLM service, users can easily translate natural language queries into SQL.

Examples

• Query: "What are our total streams for each Tom Hanks movie this month?"
  • Response: Resulting SQL and the answer.

Setting Up

1. Choose LLM: Connect to OpenAI, OCI Generative AI, etc.
2. Create AI Profile: Define schemas, tables, and connect to LLM.

Benefits

• Simplicity: Less need for detailed SQL queries.
• Future-Proof: Choose from various LLM providers.
• Security: No data leakage from LLM usage.

Summary

Generative AI models are significantly effective in improving Oracle services. They enhance data management operations without excessive human involvement. By leveraging services like AHF, the Observability and Management Platform, and Select AI, Oracle is making data handling and querying more user-friendly and efficient.

Keywords

• Autonomous Health Framework (AHF)
• Generative AI (GenAI)
• Observability and Management Platform
• Oracle Autonomous Database
• Select AI
• Oracle Infrastructure
• Large Language Model (LLM)
• Data Management
• AI-based Queries
• SQL Generation

FAQ

What is the Autonomous Health Framework? The Autonomous Health Framework (AHF) is a set of tools for monitoring and diagnosing the health of Oracle databases using AI.

How does Select AI work? Select AI translates natural language queries into SQL queries using an AI profile that integrates with a large language model (LLM).

What kind of questions can Observability and Management Platform answer? The platform can answer queries related to infrastructure health, AI workloads monitoring, and performance insights.

Can I use my own LLM with Select AI? Yes, Select AI supports multiple LLM providers like OpenAI and OCI Generative AI.

Is there a risk of SQL injection with Select AI? No, Select AI does not use binding variables in a way that enables SQL injection. It generates valid SQL queries securely.

How do I avoid issues with missing data or tables in my queries with Select AI? Ensure to create comprehensive AI profiles that include all relevant tables and schemas to improve the accuracy of generated queries.